Keywords

Introduction

In the North Rift Valley region of Kenya, potato is the third most important food crop after maize and beans and contributes 13% of total potato production in Kenya (Rees et al., 1997). However, its production and yields in the region are generally low, with a yield average of 12 t ha–1 compared to a potential yield of 40 t ha–1 (Kabira et al., 2006). The low yields are largely attributed to the unavailability of certified potato seed, low soil fertility, high incidence of late blight and bacterial wilt diseases (KARI, 2003).

Bacterial wilt caused by Ralstonia solanacearum is a serious problem in potato production in the world (Adipala et al., 2001; Lemay et al., 2003) including Kenya (Kinyua et al., 2001; Muriithi et al., 2001). Bacterial wilt has been reported to occur in the North Rift region of Kenya (Anon, 2004; Kinyua et al., 2004). Kinyua et al. (2004) reported bacterial wilt incidence of 34.3 and 32.3%, with higher limits of the disease latently in tubers in Mt. Elgon and Timboroa in the North Rift region. In spite of the above, there is insufficient information available on the disease status (prevalence, incidence, yield losses and management practices by farmers) in the region. The disease causes the greatest damage to small-scale farmers who have no access to good-quality seed and are unable to practise crop rotation because of limited land (Kidanemariam et al., 1998). This study was therefore carried out to assess the prevalence and incidence of bacterial wilt disease in major potato production areas.

Materials and Methods

The survey was carried out during the short rainy (SR) season (October to November 2006) in the highlands of the North Rift Valley of Kenya to assess the incidence and prevalence of bacterial wilt. A total of 256 farmers were interviewed in the major potato-growing areas of Trans Nzoia, Uasin Gishu, Keiyo and Marakwet districts. Potato plants with bacterial wilt symptoms in farmers’ fields in Gitwamba, Trans Nzoia district, were observed, described and collected for field and laboratory diagnosis and isolation of R. solanacearum according to the procedure described by Englebrecht (1994).

Areas covered during the assessment fell within altitudes 2153–3102 m asl, latitude 0°01–1°16ʹN and longitude 34°75–35°60ʹE. The study areas were selected through the guidance of Ministry of Agriculture extension staff and secondary information from reports and previous research work on Irish potato. Two leading potato producing divisions and two to three locations were selected in each district. The study sites were selected based on the potato cropping intensity, current potato field size and crop growth stage. At least 30 potato fields in two divisions in each district with potatoes at various flowering stages were randomly selected at intervals of 1–10 km apart. In each selected potato field, a sampling area of at least 0.05 ha was considered sufficient. In each sampled field site, the number of potato rows, plants in six to eight representative rows and plants showing bacterial wilt symptoms were recorded. Bacterial wilt incidence in each farm was determined by the number of plants showing wilt symptoms expressed as a percentage of the total number of plants assessed (Zadoks and Schein, 1979). Disease prevalence was recorded as number of potato fields with bacterial wilt expressed as a percentage of the total number of fields assessed. Whenever there was doubt on the wilt symptoms, about two to three tubers were dug up from suspicious plants and a cross-section cut made using a surgical blade or knife for observation of brown ring or bacterial ooze in clear water. During the survey, a structured questionnaire was administered by enumerators to solicit information from farmers on general potato growing with emphasis on cropping systems, practices and bacterial wilt management. A geographical positioning system (GPS) unit was used to take altitude, latitude and longitude of representative potato fields selected at random in the different areas. Data were analyzed using SPSS computer package.

Results and Discussions

Bacterial Wilt Identification

The survey revealed that most farmers could distinctly identify bacterial wilt through plant wilting and tuber oozing symptoms. However, they did not clearly understand the causal agent or the modes of spread other than through seed and soil. This suggested that farmers could manage the disease if they were equipped with the appropriate skills and knowledge on bacterial wilt epidemiology and control measures. There were positive field (wilting plants, oozing tubers, bacterial streaming) and laboratory (colony morphology) observations and/or tests for the presence of R. solanacearum in potato plants and tubers. Some colonies exhibited crescent shape on SMSA as shown in Fig. 1. Smith et al. (1996), while characterizing R. solanacearum, found that race 3 biovar II is prevalent in the major potato-growing regions (highlands) of Kenya, where races 1 and 2 were absent. Priou and Aley (1999) also reported that R. solanacearum race 3 (biovar II) strains are the most common in higher elevations of the tropics up to 3400 m asl. Thus, the bacterial wilt pathogen identified in North Rift region during this study was considered to be race 3 biovar II.

Fig. 1
figure 46_1_176274_1_En

Crescent-shaped colonies of Ralstonia solanacearum as observed on semi-selective media

Full size image

Bacterial Wilt Prevalence

Bacterial wilt occurred in 79% of all farms visited. Prevalence of bacterial wilt was highest in Keiyo district (99%) followed by Uasin Gishu (95%) and Trans Nzoia (89%) and then lowest in Marakwet district (35%). This indicated that the disease was widespread in the North Rift region of Kenya. The high bacterial wilt prevalence in Keiyo, Uasin Gishu and Trans Nzoia districts was attributed to the time the disease found its way into these districts between early 1980s and 1990s, compared to Marakwet district, where it was purportedly introduced in the late 1990s to mid-2000s. This was suspected to have resulted from uncertified seed of new varieties such as Tigoni, which were acquired from other districts by some farmers. The occurrence of bacterial wilt in Marakwet district was recorded only in Kapcherop division, whereas in areas of Koisungur (<2815 m asl) in the same division, farmers indicated that bacterial wilt was new, having been introduced in 2004 through infected seed of the variety Tigoni acquired from other potato growers in neighbouring West Pokot district. These findings suggested that in the absence of knowledge and appropriate control, diseases could spread rapidly into new areas. The use of infected potato seed and accumulation of R. solanacearum in the soil over time may contribute to the spread of bacterial wilt (Ateka et al., 2001). There was no bacterial wilt in Lelan and Kipyego areas in Kipyego division of the same district. The large distance separating these areas from other potato-growing districts, low temperatures as influenced by high altitude (above 2800 m asl) and the restricted access by road, which made these areas somehow “quarantined”, could have contributed immensely to the absence of the disease. The occurrence of bacterial wilt in Marakwet district was low or none at altitudes above 2800 m asl, whereas all farms assessed in Uasin Gishu district within the same altitude range had the disease. This implied that the cool high altitudes were not a sufficient condition to eliminate bacterial wilt, but other bacterial wilt preventive measures were also required.

Bacterial Wilt Incidence and Spread

There was a significant difference (p ≤ 0.05) in bacterial wilt incidence among the four districts, with Trans Nzoia (7.2%) and Marakwet (1.5%) districts giving the highest and lowest disease incidences, respectively (Fig. 2). This was lower than that reported by Kinyua et al. (2004) in two other areas of the North Rift region: Timboroa (32.3%) and Mt. Elgon (34.3%). Bacterial wilt incidence was significantly different (p ≤ 0.03) and had an inverse correlation (r = –0.35) with altitude (Fig. 3). This was in agreement with the findings of Ateka et al. (2001), who while studying in eastern and central Kenyan districts found a decline of 51% in bacterial wilt incidence with rise in altitude from 1800 to 2700 m asl. The disease develops slowly when the soil temperature is lower than 20°C (Martin and French, 1985; Muriithi et al., 2001). The survey indicated that bacterial wilt could be severe at altitudes below 2500 m asl, and farmers could experience higher crop loss with use of infected seed or fields at such altitudes than at mid- or higher altitudes. Average annual temperatures at the altitudes above 2800 m asl in the North Rift Valley range from 2 to 10°C (Jaetzold and Schmidt, 1983). Bacterial wilt incidence can be high in warm areas, moderate in mild areas and low in cool regions (French et al., 1996). Divisions in Keiyo and Trans Nzoia districts were not different in bacterial wilt incidence, possibly since they had near similar conditions. The significant difference in bacterial wilt incidence among districts was attributed to temperatures as influenced by altitude, cropping systems, farming practices and location in relation to other potato-growing areas. The high bacterial wilt incidence in Trans Nzoia, Uasin Gishu and Keiyo districts was attributed to contaminated soils due to continuous potato cultivation on same land over time, inadequate crop rotation intervals, use of infected seed due to lack of certified seed and poor bacterial wilt management practices.

Fig. 2
figure 46_2_176274_1_En

Mean bacterial wilt incidence in four districts of the North Rift region of Kenya, October to December 2006. Bars represent standard error

Full size image
Fig. 3
figure 46_3_176274_1_En

Influence of altitude on bacterial wilt incidence in the North Rift region of Kenya during the short rainy season, October to December 2006. Bars represent standard error

Full size image

Majority of the farmers (66%) interviewed indicated that bacterial wilt incidence was high during the short rainy season, while 34% said that it was high in the long rainy season. They further indicated that the disease initially entered their farms or spread into other farms mainly through seed potato and soils but it was also caused through black ants or rainwater flow (Table 1). Contaminated planting materials were the major source of inoculum in Kenya (Smith et al., 1996). The indication that black ants spread bacterial wilt could be attributed to their association by possibly feeding on the exudates from infected tubers, which may or may not spread the disease. The role of fertilizer in bacterial wilt spread was a perception that farmers had which could not be clearly understood, but may require further investigation. Generally the farmers indicated that they experienced bacterial wilt problem yearly (67%), occasionally (17%) and sometimes depending on the seed (13%) or field (3%). The yearly bacterial wilt incidence experienced by farmers implied that most soils were contaminated with R. solanacearum or most of the seed available to farmers was infected. The high incidence in the short rainy season could be attributed to inoculum build-up in soil from infected potato crop in the long rainy season since farmers planted at least two potato crops in 1 year and rarely removed infected and/or crop residues and tubers from the field. Warmer temperatures in the short rainy season also could have led to increased bacterial wilt incidence, which subsequently lead to yield loss. Potato yield losses due to bacterial wilt were <10, 10–25, 25–50 and >50% as perceived by 49, 30, 14 and 7% of the respondents, respectively.

Table 1 Farmers’ views on initial source of bacterial wilt in their farms and modes of spread to other farms in the North Rift region, Kenya
Full size table

Potato and Bacterial Wilt Management Practices

Potato was indicated as the second most important crop enterprise after maize, but was the leading crop at altitudes above 2800 m asl, where temperatures were too low for maize production. The majority of the farmers (63%) practised monocropping compared to 26% who practised mixed cropping. A low but positive correlation (r = 0.16) between bacterial wilt incidence and monocropping system as practised by farmers was an indication of perpetuation of the R. solanacearum in field soils, which increased disease incidence on succeeding potato crops. However, a negative correlation (r = –0.25) between mixed cropping and bacterial wilt incidence suggested that mixed cropping system reduced bacterial wilt incidence and could be a useful practice for incorporation in integrated disease management. The crops used in the mixed cropping systems included peas, beans, spring onions and maize depending on the area and/community, with peas being the dominant companion crop. In Burundi, intercropping potatoes with beans resulted in less spread of R. solanacearum than with maize intercrop (French, 1994). Within-row intercropping of potato with maize or cowpea served to contain the spread of bacterial wilt from diseased to healthy plants when initial infection was caused by either latent infection of tubers or soil infestation in Mindanao, Philippines (French, 1994). The results suggested that farmers should be encouraged to have mixed cropping as this will not only ameliorate bacterial disease but also diversify diets and income. Farmers indicated that peas at low populations was preferred to other crops as a companion in mixed cropping system because the yields of either crop were not affected negatively, possibly due to the light foliage and height advantage of peas over such crops as beans which could be smothered by potatoes.

Generally farmers indicated that they practised crop rotation (78.4%) in potato growing, but of these 55% used it as a bacterial wilt control measure (Table 2). Farmers who did not practise crop rotation either had small land holdings (41%) or lacked knowledge on its importance in bacterial wilt control (59%). Crops used in rotation with potatoes included maize, cabbage, kales, beans, oats, peas, wheat and spring onions. Other farmers used crop rotation for improvement of soil fertility.

Table 2 Bacterial wilt control methods practised by potato growers in the North Rift districts of Kenya, 2006
Full size table

The majority of the farmers practised crop rotation programme at intervals of less than 1 (44%) to 2 (37%) years, which was insufficient to get rid of the R. solanacearum in the soil before returning a potato crop in the same field. Lemaga et al. (2001), working in Uganda, reported that planting different crops in two consecutive seasons reduced bacterial wilt compared to planting the same crop and that millet and sweet potato in the rotation gave the best results. They also reported that rotation of potatoes with maize, beans, onions carrots, millet, peas and sweet potato gave significantly higher tuber yields than potato monoculture.

Retention of volunteer potato plants was a common practice according to 77.5% of the respondents. The retention of volunteers suggested that such plants carried over the bacterial wilt pathogen from one season to another through infected plants and soils. Volunteers are unsuitable in potato cropping systems because they harbour the disease and make rotation ineffective, and efforts to use clean seed are negated if volunteers are present (Tusiime and Adipala, 2000). Survival of R. solanacearum is fostered by the presence of volunteer plants from tubers of the previous crops or voluntary carriers (French et al., 1996).

Farmers who uprooted infected plants (65%) indicated that they leave such plant materials in the field (Fig. 4). The uprooting of bacterial wilt-infected plants indicated positive efforts in the management of the disease. However, poor disposal of the same and possibly timing of the operation also aggravated the spread of the disease, since 58% of the farmers indicated that they left uprooted plants and tubers in the field, 29% left along the hedges of field, 7% sometimes fed to livestock and 6% buried or burned. It was also evident through field observations that farmers left infected tubers and plant debris in the field after harvest and even when a rotational crop has been planted (Fig. 5a, b). These practices could allow bacterial wilt pathogen to spread through runoff, human and livestock movement. Potato crop residues should be collected at harvest and given to animals, and rotten (bacterial wilt-infected) tubers buried away from waterways (Priou and Aley, 1999). The farmers’ practices suggested a vicious cycle in the management of bacterial wilt, thus suggesting that education of farmers was of paramount importance in the reduction and eradication of the pathogen. Only 10% of the farmers had received some form of training on bacterial wilt management in the four districts.

Fig. 4
figure 46_4_176274_1_En

Infected potato plants left in the field after uprooting

Full size image
Fig. 5
figure 46_5_176274_1_En

Uncollected infected potato tubers in a (a) recently harvested field and (b) field planted with cabbage as a rotation crop after potatoes

Full size image

Among the respondents, 44% took more than four cropping seasons before changing or replacing their seed potato for planting, while others took one (24%), two (23%) and three (9%) cropping seasons. Replacement was shorter areas with high bacterial wilt prevalence. Most farmers (72%) selected their own saved seed potato after harvest, and 26% selected both before and after harvest. This implied that the R. solanacearum inoculum from infected tubers in such farms persisted and inevitably spread to neighbouring farms through seed exchange, soil on tools and runoff water during rainy seasons.

Selection of seed mainly after harvest suggested that farmers unknowingly mixed tubers from healthy and diseased plants, which may not be differentiated due to latent infection. This could be avoided through observation of healthy status of individual plants while actively growing. Selecting the best looking plants (positive selection) in a potato field as source of seed for the next season reduces the incidence and spread of bacterial wilt (CIP; 2007). None of the farmers interviewed had planted certified seed in the current season or in the previous 3 years. Farmers obtained their seed potato from various sources, but the most common were their own farms (47%) and neighbours (43%) (Table 3). The practice encouraged the spread of bacterial wilt within and across farms. These results supported the findings of Ateka et al. (2001) from survey studies in eastern and central Kenyan districts.

Table 3 Various sources of seed potato used by farmers in the North Rift Valley districts of Kenya, 2006
Full size table

Conclusions

While there is widespread occurrence of bacterial wilt in potato-growing areas of the North Rift Valley districts of Kenya, there are some areas that are safe from the disease in Marakwet district. These are at altitudes above 2800 m asl, such as Lelan and Kipyego. The spread and incidence of the disease were greatly influenced by low altitude, indiscriminate use of seed tubers by and among farmers due to lack of certified seed and lack of technical knowledge on bacterial wilt management and general crop practices. There exist opportunities for bacterial wilt management, which have to be utilized. Bacterial wilt in general was spreading fast particularly through infected seed, threatening the currently safe potato production areas if immediate interventions are not taken both through control and preventive or legislative measures. The findings of the survey suggest that education of farmers was of paramount importance in curbing the bacterial wilt menace in the North Rift region of Kenya security in the North Rift Valley region of Kenya. Therefore, allocation of resources towards bacterial wilt management and prevention is critical to ensure continued potato production for food. Also, the crop losses due to bacterial wilt need to be quantified across altitudes and cropping systems so as to ascertain the magnitude of the problem. This could be done alongside studies on cumulative incidence of the disease. Crop rotation studies with crops such as maize, wheat, oats, cabbage and spring onions common in specific areas among others need to be investigated. The studies will need to include a series of crop rotation intervals in those farms whose sizes can allow the practice. Where farms are too small for crop rotation to be practised, alternative bacterial wilt management options will need to be investigated.