African Crop Science Journal African Crop Science Society ISSN: 1021-9730 EISSN: 2072-6589 Vol. 5, Num. 4, 1997, pp. 395-412

POTATO IN SOUTHWEST UGANDA: THREATS TO SUSTAINABLE PRODUCTION

J. W. LOW

International Food Policy Research Institute, 1200 17th St. N.W., Washington, D.C. 20036, U.S.A.

(Received 15 August, 1997; accepted 28 October, 1997)

Code Number:CS97045
Sizes of Files:
      Text: 61.5K
      Graphics: Line drawings and tables (gif) - 70.6K

ABSTRACT

Potatoes (Solanum tuberosum L.) have emerged as an important source of cash and food for home consumption for smallholder farmers in Kabale District, southwest Uganda, since the re-initiation of breeding activities in 1989 at Kalengyere Highland Crops Research Centre. A survey conducted in 1995/96 among 134 smallholder households found mean potato yields ranging from 7.7 t ha^-1 in swampy areas to 29.5 t ha^-1 under more favourable altitude hillside sites. High demand for potatoes in Kampala, coupled with the existence of improved material, is driving the intensification of fungicide and hired labour use on potato plots and the expansion of area under potato production, albeit at the expense of proper crop rotation. However, increased disease incidence, limited availability of high quality seed, and lack of farmer knowledge on proper agronomic practices threaten the sustainability of current potato yields. In the short-term, the policy of selling highly subsidised seed to a few contact farmers should be abandoned, with increased emphasis on improved informal seed distribution systems among farmers. Moreover, significant investment is urgently needed in educating potato farmers on appropriate agronomic practices.

Key Words: Farmer knowledge, seed, Solanum tuberosum, sustainability

Resume

La pomme de terre se revele etre une source importante de revenus et d'alimentation des petits exploitants agricoles du department de Kabale, du sud-ouest de l'Ouganda, et ce depuis le re-demarrage des activites de creation de nouvelles varietes, en 1989, a la station de recherche departementale de Kalengyere. Selon une enquete realisee en 1995-96 aupres de 134 menages agricoles, rendement moyen de pomme de terre va de 7,7 tonnes par hectare dans les zones marecageuses a 29,5 tonnes/ha sur des terres a flanc de colline, plus propices et a plus haute altitude. La demande importante de pommes de terre a Kampala, capitale eloignee, conjuguee a l'existence de tubercules ameliores, entraine l'intensification de l'application de fongicides et de l'utilisation d'une main-d'oeuvre journaliere sur les parcelles agricoles de culture de la pomme de terre, ainsi que l'expansion de la superficie ocupee par cette derniere, mais au depens des techniques idoines d'assolement. Toutefois, une morbidite accrue, une disponibilite restreinte de semences de haute qualite et des connaissances lacunaires des agriculteurs en matiere de pratiques agronomiques appropriees menacent la perennite des rendements actuels de pomme de teere. A court terme, il conviendrait d'abandonner la politique consistant a vendre, a quelques exploitants de contact, des semences a des prix fortement subventionnes, est souligner davantage des systemes informels, ameliores, de distribution des semences par les agriculteurs. En outre, un investissement important est necessaire, de toute urgence, pour former les producteurs de pommes de terre aux pratiques agronomiques appropriees.

Mots Cles: pomme de terre, perennite, semences, connaissances agricoles

INTRODUCTION

The highlands of eastern Africa and Ethiopia support many of the most productive agricultural systems in the region (Stone, 1992) and include potential sites for the cultivation of potatoes (Solanum tuberosum L.). The cool temperatures needed for producing seed potato occur above 2100 metres altitude. These highland environ-ments also support some of the highest rural population densities in Africa and the smallest per capita landholding sizes (Stone, 1992).

The highland area of southwestern Uganda, Kabale District, produce the majority of Uganda's potatoes. The hilly landscape in Kabale District is interlaced with narrow and broad valleys. Altitudes range from 1,400-2,500 m, with the Kalengyere Highlands Crop Research Centre situated at 2,500 m. Annual rainfall ranges from 1,000-1,500 mm and occurs in two peaks, with high inter-year variability. The first season from mid-February-May is referred to as the short rains and the second season from September to December as the long rains. The mean annual maximum temperature is below 22.5 C and the mean annual minimum below 10.0 C (Yost and Eswaran, 1990). Kabale District is renowned for high population densities, (ranging from 200 to over 400 persons km^2 in non-forested areas), small, highly fragmented landholdings, declining soil fertility and high rates of male out-migration in search of employment. Refer to Farley (1996) for an extensive review of literature on the district.

Introduced by colonial administrators early in the 1900s, potato rapidly spread in the highland areas of Uganda as a garden crop, but was practically wiped out by late blight (Phytophthora infestans De Bary) in 1946 (Akimanzi, 1982). Imports of potato seed from Kenya re-established the crop, but yields slowly declined due to lack of suitable varieties and disease problems. In 1968/69, the Potato Improvement Programme for Uganda was initiated and by 1973, a number of varieties were released, of which Uganda Rutuku, Bufumbira, Malirahinda, Rosita, and Makerere are among those still being grown today. Potato production in Uganda may have reached 90,000 hectares in the 1970's, but the outbreak of a long period of political turmoil resulted in a dramatic drop in production, to approximately 19,000 t in 1986 (Van der Zagg, 1994).

Breeding activities were re-initiated in 1989 through a collaborative project between the International Potato Centre (CIP) and the National Potato Research and Development Programme. Three new varieties were released in 1991 (Victoria, Kisoro, and Kabale) and clean seed of existing popular varieties (e.g. Rutuku) were bulked and distributed in the early 1990s (Sikka, 1994).

By the mid-1990s, the potato emerged as a major cash crop for the majority of households in the district. Commercialised production occurred in large valleys during the dry season as well as at higher elevations (>1900 m) on the hillsides. However, the only available yield data was from on-station and a few on-farm trials. There was little knowledge about production practices and constraints faced by potato farmers. Therefore, this study had two main objectives. First, to obtain baseline farm level data on current potato production practices in sites representative of different agro-ecologies, and second, to assess the prospects for the sustainability and possible enhancement of current potato yields.

To address the second objective, the role of potato relative to other major crops in the area needs to be thoroughly understood and the degree of intensification or dis-intensification of input usage assessed. To date, most researchers believe that current level of agricultural production in the highland areas of eastern Africa has been maintained through expansion of cultivation into wetlands, forests, and marginal hillside areas as well as through decreasing fallow periods on existing fields, changing or eliminating certain crop rotation patterns, and changing the types of crops grown (ICRAF, 1987; Stone, 1992).

Considerable debate exists as to whether these densely-populated highland areas are heading towards irreversible environmental deterioration because population growth is outstripping any advances in agricultural production (the Malthusian argument) or, in contrast, population growth has provided the necessary stimulus for agricultural intensification to occur. Research conducted in nearby Ruhengeri, a major potato growing area in Rwanda, concluded that agricultural production had intensified over the past 20 years, in large part due to intensive investment by the government and donors in agricultural research and development projects aimed at food self-sufficiency (Ford, 1993). While facing similar agro-ecological conditions, no such intensive investments have occurred in the Kabale area. Hence, this research examines current intensification efforts and production constraints faced by semi-subsistence potato farmers receiving limited to negligible outside assistance.

MATERIALS AND METHODS

The research was undertaken in three phases. During the site selection visit in February 1995, it became clear that the varieties released in 1991 had only penetrated the most commercialised potato growing areas in the District. To obtain a representative sample of these varieties as part of the study, final site selection excluded remote locations. Since cash crop oriented areas are most likely to intensify production and equally likely to be the first to experience any negative consequences of such intensification, these selection criteria are logical for the objectives of the study. While many of the conditions faced by farmers in the sites are representative of other areas in the district, the quantities of different varieties grown certainly are not. Therefore, results from this study should not be extrapolated as representative of the entire Kabale District.

Subsequent field work using structured household surveys was done in the remaining two phases, September 1995 and February 1996. These periods corresponded to harvest times for potato, so that crop cuts could be obtained and accurate yield figures calculated under farmer-managed conditions. A total of 134 potato farmers were interviewed who were in the process of, or had very recently completed, harvesting a plot of potatoes. Thirty-one Kicumbi farmers, all harvesting potatoes, were interviewed in September 1995. Interviews in the other two sites, Bukinda and Kalengyere sites, were conducted in February 1996. The interview was typically conducted next to the potato plot itself to minimise disrupting the farmer's work plan and assure privacy.

To assess yields, plots currently being harvested were measured, severity of late blight assessed, and a 2 X 2 meter representative section was harvested within each plot. The sample of potatoes was weighed and tubers evaluated for marketability. In total, 140 plots were measured and 122 representative samples were weighed. In the remaining 18 cases, mostly in Bukinda, the potato field had been harvested within the previous two weeks so recall information concerning the total amount of potato harvested was considered reliable.

Late blight severity was assessed as 1 = no disease to 3 = severe disease, i.e., with severe being defined as over 50% of the plot showing the disease symptom. Using this definition, a plant with one leaf showing late blight symptoms carried the same weight as a plant with the majority of leaves affected by late blight. Hence, severity of late blight is probably overestimated in this sample, while the assessment of bacterial wilt accurately reflects the extent of the disease problem.

Approximately twice as many households were interviewed in Kalengyere than in each of the other sites, as Kalengyere is more representative of major potato growing areas in Kabale.

RESULTS

Site characteristics. Major characteristics of the three selected areas are described below.

Kalengyere. This high altitude area (2100-2400 m) of intensively cultivated, steeply sloped hills is a few kilometers from Kalengyere Highland Crops Research Centre, where late blight tolerant varieties are bred. Areas surrounding the station are among the most suitable in the district for potato cultivation. Potato commercialisation is high for both hillside and non-swampy valley bottom production, with most of the production being sold to truckers by farmers or brokers. Use of fungicides is widespread, but fungicides are not necessarily applied in recommended amounts. One farmer stated that without fungicides it would be impossible to grow potatoes in this area. Most farmers purchase chemicals from local traders supplied either from Kabale or Rwanda. Men are highly involved in commercial potato production, in contrast to other parts of the district where women are principally responsible.

Harvest time distribution in Kalengyere reflects the high potential for potato production at elevations above 2100 m and farmers appear to have parts of their hillside and valley land in potato production year-round. Wet valley land is much less available in this area than in the other two sites, and the bulk of harvesting from valley land bottom land occurs in August through November. In this highly commercialised area, the percent of harvest sold never falls below 50% in any given month. Overall, Kalengyere farmers retain 29% of their harvest for seed, and consume only 12%, the lowest percentage among the three sites. Although Kalengyere is the most distant of the 3 sites to Kabale town (50 kms), it has access to the all season Kabale-Kisoro (the next major urban centre) road.

Bukinda. This highly commercialised valley (Nyaruhanga) with its adjacent hillsides borders the main Kabale-Kampala road, some 30 kms north-east from Kabale. Bukinda is lower than Kalengyere (1700-1800 m); its warmer conditions are less favourable for potato production. However, potatoes are widely grown on both the hillsides and in the valley. The Rurangara swamp was drained in 1948 and the valley has long been utilised for vegetable production. Use of fungicides on potatoes is widespread, although, as in all areas visited, fertilizer use is non-existent. Most often, potatoes are sold to traders who arrange for a truck to come to collect produce at farmers' homes. Alternatively, smaller quantities of potatoes are sold at the local market at Bukinda village.

There are three distinct potato seasons in Bukinda, which exploit the entire range of potential production sites: September-February on the hillsides, May-November in the wet valleys, and December-May in the drier parts of the valley. Potato prices peak in March-May when the supply of high quality potatoes from higher altitude areas diminishes. Overall, Bukinda households sell a significantly greater percentage of their potatoes (60%) than do Kicumbi households (43%), placing it on par with Kalengyere in terms of degree of commercialisation. The amount of potato retained as seed is identical to that seen in Kicumbi (23%), reflecting the lower amounts of potato consumed by Bukinda households (17%) than Kicumbi households (34%).

Kicumbi. Kicumbi data represent the cultivation of wet valleys. This site is intermediate in altitude (1800-2000). Its most notable feature is the huge, flat Katuna valley. Parts of the valley floor area dominated by large fenced pastures for dairy cattle; in other parts, potato is intensively grown during the dry season; in other sections extensive sweetpotato ridges dominate. Women are the dominant producers of potato.

Potato production is highly commercialised with a major tarmac road running the length of the valley from Kabale to the Rwandan border. There is one dominant period of wet valley cultivation, planting in March-May and harvesting in late July through September. Wet conditions require extensive drainage channels to prevent water logging and destruction of the crop due to flooding. The valley is periodically flooded in March or April, which occasionally severely reduces production. Half of the harvested valley crop is sold. Hillside potato production during other times of the year does occur in some Kicumbi households, but this is mostly to obtain seed for valley planting and ware for home consumption. Kicumbi households are the least commercialised of the three areas, with no potatoes being sold from 31% of the plots harvested, compared to 11% for Bukinda and 9% for Kalengyere.

Potatoes produced in the wet valley are widely recognised to be of poorer quality than hillside potatoes. This is reflected in their lower mean price per kilogram (88 Ush) than in the other two sites,115 in Bukinda and 102 Ush in Kalengyere (1US$ = 960 Ug. sh). Wet valley potatoes do not store well and hence, are rarely kept for seed.

Household characteristics. Households across all sites were remarkably similar in their overall size and age structure. The average household had eight persons, with a dependency ratio of 0.44. The mean age of the man was 42 years and that of the woman 36 years. There were no significant differences in these characteristics across sites.

Given its proximity to Kabale town, it is not surprising that 26% of Kicumbi households had at least one household member running their own business, most typically petty trading. However, the highest percentage of households with a person earning a regular salary was in Bukinda (22%). Women in Kicumbi households were much more likely to be on their own (19% of households with no principal male) than were women in the other two areas (5% of households with no principal male). This in part explains their lower level of possession of agricultural tools and livestock than in Bukinda and Kalengyere. Livestock, especially cattle, are highly prized throughout the district and households in Kalengyere generally possessed more livestock than in Bukinda and Kicumbi.

Potato in the food system. There are seven major food crops which provide the bulk of calories for Kabale households: two grain crops (sorghum, maize); three non-grain starchy staples (potatoes, sweetpotatoes, and matoke (banana); 2 legumes (beans, field peas). Farmers view these crops as essential for food security, which, in the Kabale context, is interpreted as the ability to have sufficient food for home consumption.

Overall, the three most important crops for assuring household food security are beans, sweetpotatoes, and field peas. However, there is considerable variation in the relative importance of these crops between sites. Beans are considered the most important food security crop in Kicumbi and Kalengyere, but potatoes emerge as the most important food security crop among a third of Bukinda and Kalengyere households. Field peas are considered more important in Kalengyere than Bukinda and Kicumbi, while the opposite is true for sweetpotatoes. Approximately half of Bukinda and Kicumbi households rank sweetpotato among their top two food security crops compared to 17% of Kalengyere households.

The relative importance of the main carbohydrate sources in the diet varies seasonally, in terms of general availability and frequency of consumption. Sorghum, maize, and potato consumption is quite seasonal, while over 90% of households consume sweetpotatoes and matoke (banana) throughout the year. Seasonal patterns of potato consumption vary by area, reflecting differences in the potential for year-round cultivation in the different sites. Kicumbi farmers have one major growing season in the valley, and, consumption of potatoes by over 80% of the households only occurs in 4 out of 12 months. In contrast, in the remaining two sites, over 80% of households consume potatoes in 8 out of 12 months. The key period for purchasing potatoes occurs in March through June. Purchase for eating occurs among a larger percentage of Kicumbi and Bukinda households (34% and 30%, respectively) than in Kalengyere (14%).

Non-food cash crop production is minimal in Kabale District. The two dominant cash crops for both women and men in all sites are potato and sorghum, with beans a distant third. There are two distinct production systems, with women dominating crop production in the valley of Kicumbi, while men are heavily involved in cash crop production in Bukinda and Kalengyere. Joint production between men and women occurs in the latter two areas, thereby explaining why almost half the women report no independent cash crop sales in those two areas. The higher percentage of women in Kicumbi engaged in cash crop production also reflects the higher percentage of households with no principal man (e.g. husband, father) in the Kicumbi households (19%) than in the other two areas (5%).

Potato is by far the most important cash crop for men in all sites who are involved in crop production (Table 1). For women with separate plots, potatoes also dominate in Kicumbi and Bukinda and are second to sorghum in importance in Kalengyere. Sorghum is second in importance for men in Bukinda and Kalengyere, and for women in Kicumbi and Bukinda.

The dominance of potato as a revenue source in Bukinda and Kalengyere is particularly evident when one looks at the percent of total gross cash sales from crops derived from potato (Table 1). If one combines crop sales of both men and women, 73% of total revenue in Bukinda and 76% of revenue in Kalengyere derives from potato sales. Even in Kicumbi, where the total cash earned from crop sales is over 50% lower than that for households in the other two areas, potato sales account for 48% of total cash crop revenue. Men dominate the revenue derived from potato sales in Bukinda and Kalengyere, while women control the majority of cash from selling potatoes in Kicumbi.

Over half of the respondents reported that potato was increasing in importance as a source of cash and as a food for home consumption. The highest percentage increase of potato as a source of cash over the past five years occurs in Kalengyere (62%), followed by Bukinda (51%) and Kicumbi (48%); the increase is due to increasing market demand and the high productivity of the crop, in part due to new varieties. Increase in home consumption was attributed to increasing family size and a changing food preference. Potatoes as a food are becoming more popular and greater home consumption is in part a reflection of greater availability at the household level as a by-product of planting more of the crop for cash sale.

In spite of the better market for potatoes, a high percentage of farmers (38%) reported a decline in the importance of potato as a cash source during the past five years. Bukinda and Kicumbi growers cited higher percentages of decline (44% and 38%, respectively) than did Kalengyere growers (33%). The overwhelming reason behind this trend was declining yields, with some farmers specifically noting that the decline in yields was due to poor soil fertility or disease attack. Yield decline or reduced production was also the major reason behind the 26% of growers noting a decreasing importance of potato as food for home consumption.

Dominant potato varieties and their utilisation. The characteristics of the main potato varieties released by the National Potato Programme are summarised in Table 2. Three important varieties found in this survey are briefly discussed.

Uganda Rutuku, also known as Rusina or Otankubura, is the most demanded variety in Kabale because of its high marketability. It is preferred for chip production in Kampala, and its red skin and cream yellow flesh colour have become the preferred colours in Kabale, demanding a higher price than any white-skinned varieties irrespective of flesh colour.

Victoria has been the most adopted of the three varieties released in September, 1991. It possesses the highly desired red skin and cream flesh colour and has a shorter maturity and dormancy period than does Rutuku. It is better adapted to lower elevations (1200-1750 m) than Rutuku but farmers report that it is more susceptible to late blight than Rutuku. The demand for Victoria seed is high and seed is in short supply even in the three highly commercialised research sites. Thus, even the price of ware Victoria is significantly higher than that for Rutuku. In February 1995, Victoria ware potato was selling for 190 Ush per kilogram, compared to 139 for Rutuku.

Cruza has a starchy texture, fast cooking time, and tolerance to late blight and bacterial wilt attack, but it has distinctive purple marks on the skin, and purple rings throughout the white flesh on the inside, making it highly undesirable to buyers from Kampala. Cruza can only be sold to Kampala buyers during April-June when potatoes are in short supply. Otherwise, Cruza mostly sold locally at significantly lower price than red-skinned varieties. However, Cruza has a niche among poorer households in the area as an important food security crop as farmers at lower altitudes who cannot afford fungicides plant Cruza for home consumption.

Farmers in this sample have had considerable experience growing potatoes. Only 18% have been growing potatoes less than 6 years, while another 18% have grown it for over 25 years. On average, Kabale farmers have cultivated potatoes for 14.8 years.

Varietal turnover is not fast, as farmers retain much of their own seed for replanting. During the past five years, 67% adopted only 1 or 2 new varieties, and 62% dropped 1 or 2 varieties.

In addition to ease of selling, four important reasons mentioned for varietal adoption were high yield, good tuber size, late blight resistance, and interest to try the new variety. Early maturity is also a highly desirable characteristic.

Cruza was the variety most likely to be dropped during the past five years by commercially conscience Bukinda and Kalengyere farmers. But, only 10% of Kicumbi farmers dropped Cruza, compared to 57% of Bukinda and 44% of Kalengyere growers. While some farmers appreciated Cruza's quick cooking time, its tendency to crumble when over-cooked is disliked by others.

The overall pattern in this sample is to eat a fifth, retain a fourth (for seed), and sell slightly more than half of harvested potatoes. Price variability and differences in the percent of harvest sold in each area reflects the types of potato varieties grown as well as their yields. In general, hillside potatoes command a higher price than valley potatoes; red-skinned potatoes command a higher price than white-skinned potatoes. Cruza, a white-skinned variety with purple splotches, commanded a significantly lower price than other varieties and was more likely to be consumed, than kept for seed or sold.

Current yields and their determinants. Mean yields differed significantly among the sites (Table 3). The high mean yield of 29.5 t ha^-1 among Kalengyere farmers is similar to yields obtained at the research station. While yields per hectare are significantly lower in Bukinda (19.6 t ha^-1) than in Kalengyere, yield per plant and marketable yield per plant do not differ significantly. The lower yield per hectare is associated with the lower planting densities in Bukinda (39,554 plants per hectare versus 50,207 plants per hectare in Kalengyere). The agricultural extension service recommends 44,000 plants per hectare, so Kalengyere farmers are planting at a somewhat higher than recommended rate, while Bukinda farmers are planting somewhat less. The lower planting densities in Bukinda reflect the higher percentage of plots being intercropped in Bukinda than in Kalengyere. In Bukinda 59% of the potato plots were intercropped with maize, compared to only 9% of plots in Kalengyere, with bananas being the preferred intercrop (Table 4).

Yields in Kicumbi are one-quarter of those found in Kalengyere, 7.7 tons per hectare. In part, this is due to lower planting densities (22,418 plants per hectare) because of the extensive ridges constructed for proper water drainage in the swamp and 36% of plots being intercropped with beans. Also, yield per plant (0.35 kg) is 69% of the yields per plant in the other two sites (0.51 kg). The two driving forces behind the low yields in Kicumbi were viral diseases, with 46% of plots being severely affected, and bacterial wilt, with 30% of the plots being severely affected (Table 4). Since aphids are associated with virus transmission, it is not surprising that aphids were found on 82% of the Kicumbi plots. Presence of aphids and virus problems were much reduced in Bukinda and Kalengyere.

Evidence of severe cases of late and early blight (Alternaria solani), were much higher in Bukinda and Kalengyere than in Kicumbi (Table 4). Late blight usually attacks earlier in the season than early blight and the vast majority of farmers spray fungicides in response to these diseases. In contrast, there was very little evidence of late blight in Kicumbi, most likely due to the swamp crop being planted towards the end of the rainy season and growing during the driest time of the year.

Early blight attacks towards the end of the season and tends to have a much less destructive effect on overall production than late blight. Farmers, however, do not distinguish between the two kinds of blight, both being referred to in Rukiga as okubabuka or okusya (which literally means burned-referring to the blackened leaves).

It is important to note that the mean potato plot size on these farms is very small, ranging from a low of 560 ha in Bukinda to a high of 812 ha in Kicumbi. Thus, even the high yield of 29.5 tons per hectare in Kalengyere translates into 2062 kilograms (approximately 19 bags) of potatoes from an average plot size of 699 ha.

One of the major determinants of potato yield is seed quality. In this case, the higher yields in Kalengyere may be due to the greater usage of seed that originates from hillside plots than from valley bottom or swampy sites. Seventy percent of Bukinda and 78% of Kalengyere farmers use their own seed. Kicumbi cultivators, on the other hand, are heavily dependent on market purchase of seed. Fifty-eight percent of plots planted since July 1994 in Kicumbi used seed of unknown origin in terms of type of land (hillside vs. swamp, for instance). No farmers in Kicumbi received seed from an extension service or the research station and only 3% of Bukinda plots and 5% of Kalengyere plots obtained seed from these formal sources.

Very significant differences do exist in the cost of potato used as seed for varieties in high demand (Victoria, Rutuku) and those in low demand (Cruza, Malirahinda). On average, farmers spent 28,000 Ush for a 110 kg sack of Victoria, but the price for germinated Victoria seed could reach 40,000 Ush per sack. In contrast, the mean amount spent on Cruza was 13,800 Ush. On average, farmers used the same seed longer in Bukinda and Kalengyere (4.2 and 4.6 years, respectively) than in Kicumbi (1.6 years). Therefore, continual cash outlay for seed was a regular feature of the potato production system in the swamps of Kicumbi. The number of years the same seed was used depends on how long it has been available, Victoria being in use for only 1.5 to 2 years in all sites, compared to 1.9 to 5.2 years for Rutuku.

The demand for high quality seed from the Kalengyere Highland Crops Research Centre vastly exceeded supply. It serves potato growing areas throughout the country, but produces relatively little high quality pre-basic seed. In 1995/96, the policy was for a few key contact farmers to receive seed in each district and these farmers would in turn multiply and sell seed to their neighbours. Seed is sold to these farmers at prices far below the market price of seed. On average, Rutuku was selling at 52% of its market value and Victoria at 68%. Farmers receiving the highly subsidised seed ended up selling the vast proportion of their harvest to traders from Kampala in order to meet their immediate cash needs. In group discussions with farmers in both Kalengyere and Bukinda, farmers complained bitterly that the same contact farmers received seed every year from the station and that they were not able to obtain seed for themselves in a timely manner. The current system of subsidising seed to a few farmers has created problems of jealousy among neighbours and much, high quality second and third generation seed has been consumed rather than planted.

Two diseases constitute the greatest threats to sustained potato production in Kabale, late blight and bacterial wilt (Hakiza et al., 1997; Tusiime et al., 1996a). Cool, wet conditions favourable to late blight attack can occur in eight out of twelve months in Kabale (Sikka et al., 1997). The major strategy for combating late blight is to spray fungicides.

Bacterial wilt can also cause severe losses once established, with the potato plant completely wilting when seriously infected. Tubers can be infected, however, without showing any external symptoms unless the infection is serious (Tusiime et al., 1996b). Since chemical control of wilt does not exist, integrated control measures are advocated, combining use of resistance varieties, planting disease-free tubers, crop rotation with non-host plants (avoid planting tomato, tobacco, eggplant), removal of volunteer potatoes (plants that emerge from incomplete harvesting of the previous potato crop), and minimum tillage during the growing season to avoid injury to roots (Martin and French, 1985; Tusiime et al., 1997).

To assess the extent of farmer knowledge about these diseases, open-end questions were put to farmers asking them to identify the symptoms of the disease, the causes of the disease, and the strategies they use to avoid or manage the disease in their potato plots. While farmers mentioned at least one major symptom of each disease, knowledge of causes and appropriate agronomic practices to prevent infection was limited.

In general, late blight is considered a much more serious problem than early blight by farmers. The symptoms of late blight (dry, blackened leaves; small tubers) are well-recognised. The vast majority (86%) stated that excessive rain is the cause of late blight, with 18% mentioning cooler temperatures as well. No one mentioned the very distinctive white spots on the underside of leaves associated with the fungus or other facts that would indicate knowledge of a disease pathogen.

While farmers recognised wilt, knew that wilt-infected tubers rot, and had a distinct name for it in the local language (okwoma), 38% could not cite any associated cause. Moreover, 37% of farmers take no steps to prevent either the infection or spread of the wilt. Seventeen percent did recognise that wilt was caused by infected seed and an additional quarter associated it with a variety of soil problems (planting potato after potato, loss of soil fertility, infected soil).

Almost half of farmers said they uproot wilted plants when they encounter them in their fields. However, if that figure was entirely accurate, fewer infected plants would have been encountered during the yield assessment exercise than was the actual case. Ten percent mentioned crop rotation and only 2% cited seed selection as preventive strategies.

The retention and spread of disease from potato tubers left in the ground after harvest are major sources of infection. Yet only 19% of farmers reported ever uprooting "volunteer" potato plants (technically referred to as roguing). Most farmers consider this volunteer potato crop an important source of food, with 23% claiming to get "a lot" of potatoes from the volunteer crop, 76% reporting some yield, but not much.

Crop rotation is essential for avoiding disease intensification in potato-based systems. Thus, examining the three year crop histories for the representative potato plot provides an insight as to whether farmers are practicing crop rotation in a manner that will minimise disease intensification, particularly for bacterial wilt. Ideally, potato should never be planted after potato and crops rotated with potato should not include crops susceptible to the same pathogens, such as tomatoes and eggplants.

Potato is rotated with beans, sorghum, matoke, and maize extensively, and to a lesser extent with peas, wheat, and tobacco. Peas and wheat are more important in rotations on poorer soils on the upper parts of the hillside. Only 9 cases (7%) included sweetpotato in the rotation. Four cases of rotation with tomato were found, all in Bukinda, the area most known for vegetable production. The problem with this rotation is that tomatoes are very susceptible to both late blight and bacterial wilt, two of the most important diseases of potatoes.

Potato was planted more than once in 3 years on 86% of the plots. Twenty-eight percent of plots had potato 3 times, and 3.7% (6 cases) more than 3 times in 3 years. Potato was planted for two, three and more than three consecutive seasons on 19, 10 and 3% of the plots, respectively. Moreover, on 27% of plots only a single planting of a different crop occurred in between two potato plantings.

Bukinda farmers were the least likely to plant potato directly after potato. Bacterial wilt was evident on half of the Kalengyere plots (Table 4) and yield decline is likely if current practices continue.

During the three year period since January 1993, only 4 representative potato plots in Kicumbi were sprayed, compared to 56 in Bukinda and 106 in Kalengyere. Kalengyere farmers bought their chemicals from small shops and Bukinda farmers purchased from small shops and a nearby market. Only 5% of those spraying purchased their fungicides in Kabale town. Most Kalengyere farmers went by foot to buy their chemicals, whereas half of the Bukinda farmers used bicycles or used motorised transport.

In 96% of the cases when spraying occurred, a fungicide was used to combat late blight, either preventively or in response to an outbreak of late blight. The majority of farmers spraying stated that they began spraying before late blight symptoms appear. The brand name of spray used in all but one case was Dithane (Mancozeb 80% WP), on sale in a single shop in Kabale for Ush 8,000 per kilogram in February, 1995 and Ush 11,000 per kilogram in February 1996.

Use of spray differed significantly between Kicumbi and the other two sites. When the number of sprayings were examined on the representative plots by agricultural year, over 90% of Kicumbi households did not spray at all in either 1993/94 or 1994/95. In 1995/96, almost half of Bukinda farmers sprayed 5 times or more, compared to 25% of Kalengyere farmers. Less than 20% of Bukinda and Kalengyere farmers never sprayed. On average Bukinda farmers sprayed significantly more times per plot (4.8 times) than did Kalengyere farmers (3.9 times). Farmers wear no protective clothing while spraying and children as well as adults are involved in spraying. It is also possible that farmers are not applying recommended rates.

The cost of the chemical itself comprises 80% of the total cost of spraying. Seventeen percent of total cost in both areas went towards renting sprayers and hiring labour to spray. Due to lower total costs and greater number of sprayings per plot in Bukinda than in Kalengyere, Bukinda farmers have significantly lower mean costs per round of spraying (3,974 Ush) than do Kalengyere farmers (5,277 Ush).

Farmer awareness concerning the proper size and number of stems seed should possess on the whole was good. Ninety-three percent of farmers from all three areas correctly believed that it was better to plant potatoes that had more than one emerged stem. This was surprising as informal market surveys revealed significant quantities of germinated seed being sold that were characterised by a single emerged stem or apical dominance. Farmers were also asked to select the best size of seed tuber for planting among a sample of tubers ranging from 25mm to 70 mm in diameter. Only 6% of the sample chose small seed (25 and 35 mm), and 14% chose a size too large to be economic (70 mm). The majority selected seed in the appropriate middle range, 31% selecting tubers 40 mm in diameter; 47% opting for the most desirable size of 50 mm (Van der Zagg, 1994).

Almost all farmers (96%) stored potatoes at some time during the year. Bukinda and Kalengyere farmers were more likely to keep potatoes in a separate store but Kicumbi farmers were more likely to store in the home. The station and extension service recommend that potatoes be stored on racks that are above the floor to reduce rotting. Only 12% of the 134 farmers surveyed stored their potatoes on such racks, with 10 of the 15 cases coming from Kalengyere. Seventy percent kept their potatoes in heaps directly on the floor, 16% placed them on top of dried grass or leaves, and 3% on tops of mats or stones.

DISCUSSION

This section addresses the question of the sustainability of potato cropping systems as they now exist in Kabale District. The discussion draws on survey results as well as numerous group discussions held with farmers in the area.

Results from on-farm yield assessment of potato production indicate that farmers in the high altitude areas surrounding the research station in Kalengyere are, in general, obtaining excellent yields (29.5 t ha^-1) given the current level of input use and it would be highly desirable to sustain these yields, albeit with lower production costs. At the other end of the spectrum, yields for Kicumbi valley are already low (7.7 t ha^-1) and the system is under severe threat. The question here is not of sustaining, but enhancing yields. The case of Bukinda falls somewhere in between sustaining mean yield levels of 19.6 t ha^-1 at these lower altitudes (1800-2100m) and targeting for the high yields achieved at Kalengyere. If this is done, potato would remain a very important cash crop in this area.

There are several factors that will encourage sustained potato production of potato in Kabale District. These include the:

High demand for potatoes in Kampala, the capital city. Incomes are continuing to rise in Uganda, and the demand for potatoes by both poor and wealthy urban consumers is likely to continue to rise (Bouis et al., 1996).

Improved road infrastructure. The rehabilitation of the main road from Kabale to Kampala enabled potatoes to be efficiently delivered to the capital city in the early 1990s. Indications are that this road structure is being sufficiently maintained.

Increasing entrepreneurship. Improving incomes, improving infrastructure, and a policy of market liberalisation mean that an increasing number of individuals will become involved in long-distance trading and that eventually, there will be more frequent transport available for moving crops to more distant markets. In addition, given the rising demand for chemical inputs such as fungicides, pressure exists for increasing entrepreneurship (number of traders and shops) marketing these products in Kabale District.

Increasing input use. Farmers are most willing to invest in inputs (e.g. fungicides, hired labour) and the labour to apply inputs (e.g. animal manures) on high value crops like potatoes. Given the high response of potatoes on-station to fertilizer applications (Sikka, 1994), farmers are likely to spend money on inorganic fertilizers as well, if they were available at reasonable cost.

Existence of late blight tolerant varieties. The existence of late blight tolerant varieties have enabled many of the poorest (e.g. Kicumbi swamp farmers) to produce a potato crop without the use of fungicides. Varieties highly susceptible to late blight will not survive under the socio-economic conditions of Kabale. Given the high and ever increasing cost of fungicides and their hazardous side effects the continued availability of late blight tolerant varieties is critical for the sustainability of potato production in Kabale. According to an evaluation of the pre-basic seed programme at Kalengyere Highland Crops Research Centre (Van der Zaag, 1994) the current system for producing high quality seed is sustainable at the current levels of production. The levels of production, however, are insufficient to meet the ever increasing demand for new seed. Therefore, the critical task will be maintainence of seed quality as it is multiplied and distributed by farmers.

The repeated planting of potatoes in the Kicumbi swamp explains in part the higher incidence of bacterial wilt and the much lower yields encountered in Kicumbi compared to the other locations. However, much of the repeated planting of potato occurs on favourable black loam (eririkwiragura) soils in Kalengyere. To date, mean yields have remained surprisingly high given this practice.

The lack of knowledge concerning the need to use "clean" seed and practice crop rotation and other appropriate agronomic practices does not bode well for managing late blight and bacterial wilt. Farmers in this sample are not inexperienced with potato, over half had been producing potato for over 10 years. This lack of knowledge reflects the weaknesses of institutions in the area. The number of extension staff for the entire district are few and potato agronomy has never been integrated into district level primary school education.

Factors operating against the sustained production of potato throughout the next century can be grouped into three broad categories: (1) Severe constraints to implementing appropriate agronomic practices regarding potato, (2) Weak national institutions, and (3) Increased competition from other parts of the country. Each of these factors will be discussed below.

Severe constraints to implementing appro-priate agronomic practices. The greatest threat to sustained potato production in Kabale District is the increasing incidence of late blight and bacterial wilt. There are many recommended agronomic practices that can be practiced by farmers which will reduce the rate at which these diseases spread and the severity of incidence (see for example Tusiime et al., 1997). These practices are summarised in Table 5, along with the major constraints to implementation that were identified in this study. Each of these factors will be discussed below.

Because of it's high value, farmers strive to produce potato while still having adequate amounts of other foodstuffs for home consumption. Ideally, potato should be rotated with other crops but many farmers do not recognise the value of or are constrained by limited amounts of land suitable for potato production and sufficiently close to home to avoid excessive theft. While many farmers do rent land, the high cost of land rental lowers profit margins considerably.

Farmers are also constrained by the lack of alternative crops to use in rotation that are equally as profitable as potato. Sorghum is the closest alternate, but its use is limited by planting time requirements which restrict it to one major growing season. Maize is the second, but its market value is typically half that of sorghum. Beans and peas sell for a high price, but are low yielding and often need to be reserved for home consumption. Farmers complain that wheat is not profitable. Bukinda farmers see tomatoes a profitable alternative, but this crop is susceptible to the same pathogens as potato. Sweetpotatoes command extremely low market prices and often do not yield well on soil types preferred for potato.

In spite of having considerable experience with the crop, the lack of knowledge among farmers regarding proper agronomic practices and disease management is the most serious constraint to sustained potato yields in Kabale. While farmers can recognise the symptoms of both diseases and know the association between high humidity and late blight, they know little of the causes and spread of bacterial wilt and of alternatives for management of the diseases. Roguing volunteer potato plants may encounter considerable resistance as many farmers perceive these plants as an important source of food.

While increasing use of animal manures and crop residues on potato fields is occurring, steep terrain and limited availability make manure application a labour intensive process which, encourages farmers to keep re-using plots near to home for potato cultivation. The preferred management strategy is to wait until yields have fallen to unacceptably low levels, then graze livestock on the fallow plot so that manure transport is not an issue. The lack of availability of inorganic fertilizers is a major constraint. Given Uganda's landlocked status, and the distance of Kabale from Kampala, it is unlikely that fertilizer use could be undertaken on a large-scale without being subsidised.

Another serious constraint is the lack of sufficient quantities of clean seed available to farmers. High quality seed from Kalengyere is sold to a limited number of contact farmers in the area at a subsidised rate. In theory, these farmers multiply seed and sell to other farmers in the area. In practice, most of the next harvest is marketed as ware or retained by the contact farmer for replanting and only a few other farmers end up purchasing the new varieties for seed. Furthermore, many of the same farmers have been used year after year and this has created jealousy among farmers in certain areas.

Farmers lack knowledge of seed selection and storage. The best size of potato to retain as seed is often sold to the ware potato market. Thus, many of the poorer farmers often retain smaller, less desirable potatoes for replanting and sell all potatoes of marketable size. Moreover, to ensure that the skin of the potato does not easily peel during storage, stems of the potato plant should be cut off a week or so prior to harvesting. Many farmers fear doing this, however, as cut stems advertise to potential thieves that the potatoes are ready to be harvested.

Weak national institutions. While Uganda has been going through a remarkable period of recovery since 1986, it has limited resources to invest. Many special war recovery programmes which distributed tools and seeds to re-establish agricultural production have now terminated. The agricultural extension service in Kabale District has few staff and extremely limited resources. As a consequence, the only farmers receiving extensive training on potato are the contact farmers selected for seed multiplication.

Kalengyere Highland Crops Research Centre also has limited resources and must meet the demand for seed throughout the country as well as Kabale District. To date, distribution has occurred through links between the National Potato Programme and the Ministry of Agriculture's extension service. Limited efforts have been made to involve non-governmental organisations (e.g. CARE in Kabale) in seed distribution efforts. A one-time distribution of seed was also made to a small number of women's groups in Kabale in 1993. Given its current staff and resource limitations, Kalengyere Highland Crops Research Centre is unlikely to be able to massively expand its distribution efforts.

Increased competition from other parts of the country. Potatoes coming from the Kalengyere area are widely recognised as the best quality potatoes available. A niche market for these potatoes has developed whereby they are the preferred potato for hotels selling chips (french fries). However, the successful dissemination of potato seed in the early 1990s to lower elevation areas that are closer to Kampala, and hence face substantially lower transport costs, has resulted in the development of significant competition for Kabale farmers. Kalengyere farmers in the highest elevation areas that are most difficult to reach are more adversely affected than are Bukinda and Kicumbi farmers who are closer to Kabale town. Moreover, Kicumbi farmers produce large amounts of swamp potatoes at a time of year when there is not much competition from other parts of the country.

Kampala is likely to remain the major market for potatoes in the next twenty years. There is limited scope for additional markets to develop as Kabale is adjacent to Rwanda and Zaire to the south and west, and any growth in consumption to the north and east can be met by local potato producers. Thus, the competition will dampen the growth rate in the value of a bag of harvested potatoes and the key to economic survival will be adopting recommended agronomic practices, reducing production costs, and greater emphasis on selling higher quality potatoes for use as seed than strictly for ware.

The evidence presented indicates that the factors mitigating against sustained potato production at current levels are likely to outweigh encouraging ones in the longer-term. This leads to the conclusion that without significant investments in farmer education and improved distribution of high quality seed, current yields of potato are not sustainable.

CONCLUSION

The potential for increased potato production from improved agronomic practices is high. Preliminary trials using such practices (e.g. using clean hoes, minimum disturbance during cultivation, roguing diseased plants when they first appear) and clean seed showed that yields could increase two to four times (Lemaga, personal communication, International Potato Centre, April 1996). Of course, one consequence to increased farmer knowledge concerning the importance of clean seed, is that demand for such seed will dramatically increase. This will put further pressure on a seed system that cannot meet current demand.

The need for farmers to become more highly involved in the multiplication and distribution of seed is quite evident. Participatory research should be conducted with farmers on developing successful seed multiplication systems in the East African context and learning integrated disease management techniques. Over half of the farmers in this sample were members of some kind of farmer's organisation or society. The stronger of these groups should be targeted for initial intervention.

In the short-run, the KalengyereHighland Crops Research Centre should abandon its policy of highly subsidising quality seed of new varieties, which will result in more farmers receiving seed and increase the likelihood that the seed will not be sold to the Kampala market after one growing season. Emphasis should be placed on involving different farmers annually to ensure distribution of new varieties to more remote areas of the district as well as those with easy access to transport. Involvement of non-governmental organisations or special project funding may be called for to ease the administrative burden of such tasks.

Declining soil fertility is a major constraint to crop production in the area that is likely to continue to increase in importance. In the 1950s, 1960s, and 1970s, farmers were able to respond to declining hillside soil fertility through swamp reclamation and opening fresh lands for crop production. Most swamps close to urban centres of any size have been reclaimed, and restrictions exist against draining any additional wetlands. Policy research needs to be conducted on the feasibility of importation, subsidisation, and cost-effectiveness of appropriate inorganic fertilizers as it is clear that the limited available organic manures and the labour to apply them are likely to remain significant constraints in the long-term. Farmers are only likely to be willing to pay for inorganic fertilizers for use on high-valued crops like potato, sorghum, and some vegetables (e.g. tomato).

Finally, research efforts should also concentrate on testing green manure crops adapted to higher altitudes that could be planted on fallow fields. Use of these manure crops are the best possibility for increasing the organic matter content in the long-term on depleted soils (Mugendi et al., 1997).

ACKNOWLEDGEMENTS

This research was undertaken while the author was a Rockefeller Foundation Social Science Fellow at the Nairobi Regional Office of the International Potato Centre. The assistance of William Rufumbaguza, Margaret Ruhara, Francis Alacho, Berga Lemaga, and AlineMarie Kamerwa in conducting the study is deeply appreciated. Additional logistic support was provided by Erika van Duijl, N.B. Lutaladio and Jane Wyant. Special thanks are extended to Peter Ewell, the CIP Regional Representative in Nairobi, for his interest and support during the study.

REFERENCES

Akimanzi, D.R. 1982. Potato development and transfer of technology in Uganda. In: Potato Development and Transfer of Technology in Tropical Africa. Nganga, S. (Ed.), Inter-national Potato Center, Lima, Peru.

Bouis, Howarth, E. and Scotti, G. 1996. Demand for high-value secondary crops in developing countries: The case of potatoes in Bangladesh and Pakistan. Food consumption and nutrition division discussion paper No. 14, IFPRI, Washington, D.C.

Farley, C. S. 1996. Smallholder Knowledge, Soil Resource Management, and Land Use Change in the Highlands of Southwest Uganda. Dissertation, University of Florida.

Ford, R. 1993. Marginal cropping in extreme land pressures: Ruhengeri, Rwanda. In: Population Growth and Agricultural Change in Africa. Hyden, G., Kates, R.W. and Turner II, B.L. (Eds.), pp. 145-186. Gainsville, Florida, University of Florida.

Hakiza, J.J., Kakuhenzire, R.M., Alacho, F.O. and Mukalazi, J. 1997. The international late blight and coordinated regional yield evaluation of potato genotypes in Uganda. African Crop Science Conference Proceedings 3:981-987.

ICRAF. 1987. Agroforestry Potentials for the Land-Use Systems in the Bimodal Highlands of East Africa: Uganda. AFRENA Report No. 4. Nairobi: International Center for Research on Agroforestry (ICRAF).

Martin, C. and French, E.R. 1985. Bacterial Wilt of Potato: Pseudomonas solanacearum. Technical Information Bulletin 13. International Potato Centre, Lima, Peru. 16 pp. (Second edition, revised).

Mugendi, D.N., Mochoge, B.O., Coulson, C.L., Stigter, C.J. and Sang, F.K. 1997. Effect of incorporating Cassia siamea prunnings on maize yield in an alley cropping trial in semi arid Kenya. African Crop Science Journal 5:201-207.

Sikka, L.C. 1994. Technical Assistance to the Uganda National Potato Research and Development Program, 1989-1994. End of Project Report. International Potato Centre, Nairobi, Kenya.

Sikka, L.C., Ewell, P.T. and Alacho, F. 1997. Potato seed production in developing countries: a Ugandan experience. African Crop Science Conference Proceedings 3:121-129.

Stone, P. 1992. The Status of the World's Mountains: A Global Report. London: Zed.

Tusiime, G., Adipala, E., Opio, F. and Bhagasari, A.S. 1996. Screening solanum potato genotypes for resistance to Pseudomonas solanacearum in Uganda. African Journal of Plant Protection 6:108-118.

Tusiime, G., Adipala, E., Opio, F. and Bhagasari, A.S. 1997. Towards developing an integrated control package for potato bacterial wilt in lowland Uganda. African Crop Science Conference Proceedings 3:989-994.

Van der Zagg, P. 1994. Seed Potato Production in Uganda: Present Status and Future Direction. Evaluation for the National Agricultural Research Organization (NARO) of the Government of Uganda, the International Potato Center (CIP), and the United States Agency for International Development. Unpublished report, Canada.

Yost, D. and Eswaran, H. 1990. Major Land Resource Areas in Uganda. Kampala: United States Agency for International Development (USAID).

Copyright 1997, African Crop Science Society

The following images related to this document are available:

Line drawing images