African Crop Science Journal African Crop Science Society ISSN: 1021-9730 EISSN: 2072-6589 Vol. 7, Num. 3, 1999, pp. 223-244

Editors: C.S. Gold, E.B. Karamura, A. Kiggundu, F. Bagamba and A.M.K. Abera

Editorial Secretariat, Faculty of Agriculture and Forestry, Makerere University, Uganda

The role of banana in the local diet is of greater importance in Uganda than anywhere else in the world. The East Africa Great Lakes plateau has evolved as a secondary centre of Musa diversity with at least 84 locally evolved, unique clones including cooking (matooke) and brewing (mbidde) types These endemic clones have been collectively termed East African highland banana (Musa spp., group AAA-EA).

Traditionally, highland cooking banana has been the most important staple food in the densely populated central region of Uganda from the north-central shore of Lake Victoria to the west bank of the Nile. By the end of the 19th century, the Baganda and Basoga around Lake Victoria, the Konjo and Bamba in the Rwenzori region and the Gisu near Mount Elgon were predominantly banana eaters. More recently, banana has gained an important foothold in the country's southwest.

Highland cooking banana has a number of advantages over other staple crops in the region. Its extended harvest period assures food and income sources throughout the year, while its ease of production and preparation make it even more attractive to subsistence agriculturalists. Banana also reduces soil erosion on steep slopes and serves as a source of mulch for maintaining and improving soil fertility. Until recently, highland banana has been a highly sustainable crop with stands often remaining productive for 30 years or more.

Since the 1970s, however, there has been westward movement of cooking banana within Uganda. This geographic shift has reflected the relatively recent phonomenan of loss of sustainability and production declines (i.e., lower yields and reduced stand life) in the central zone, concurrent with increased production and market opportunities in the country's southwest. However, early stages of decline are already in evidence in areas of relatively recent crop introduction (e.g. Masaka district).

Apprehension that the factors leading to the loss of sustainability in the central region may be replayed in extant production areas has raised concern about the future of cooking banana in Uganda. Therefore, a multi-disciplinary study was conducted at 9 central and 6 southwestern sites to document shifts in cooking banana production and to elucidate the causes behind these shifts.

Regional summaries and important trends underlying geographic shifts in banana in Uganda have been reported by Gold et al. (1999). This paper was, essentially, a composite of 15 rapid rural appraisals. However, the synthesis of results presented in the paper left out much of the information that was collected during intensive farmer interviews and on-farm verification studies, detailing the dynamics at each of the 15 study sites.

This includes site descriptions, relative production trends of all food and cash crops, lists of banana clonal diversity for each village, descriptions of banana stand sizes and productivity, site rankings of causes behind banana decline or crop expansion, weevil and nematode pressure and shifts in agronomic practices.

We are grateful with the editors of the African Crop Science Journal that they concur with our view in the importance of documenting these data. We also appreciate the support of the United States Agency for International Development (USAID grant LAG-G-00-94-00007-00) in providing the costs of printing this publication.

The highland green cooking banana (Musa spp. type AAA-EA) is the most important staple crop in southern Uganda and throughout the East African Great Lakes Region. The crop is a key component in both the food security and agricultural sustainability of the region. Well managed banana stands often persist for 30 or more years, even under low input conditions. At the same time, an extended harvest period ensures food and income throughout the year. Highland banana reduces soil erosion on steep slopes and are principal sources of mulch for maintaining and improving soil fertility (INIBAP, 1986). As such, banana-based cropping systems have provided the most sustainable option under East African mid- and high elevation ecological conditions. Uganda is the region's leading producer and consumer of banana.

Between 1970 and 1990, Uganda witnessed the decline of cooking banana production in traditional growing areas in the central region (Iganga, Jinja, Kamuli, Luwero, Mpigi, Mukono districts). Meanwhile, Kampala's non-Asian population rose by 750% between 1960 and 1990. The resulting increase in urban market demand for cooking banana has furthered crop expansion in the country's still productive zones. As a result of these two factors, crop production has become concentrated in the southwest. Currently, 63% of the land given over to cooking banana is found in five districts in southwestern Uganda: Mbarara (20%), Bushenyi (16%), Masaka (11%), Mubende (8%) and Rakai (8%) (Uganda Ministry of Agriculture, 1992).

Apprehension about the sustainability of banana in current production areas has been highlighted by three factors: (1) severe banana weevil (Cosmopolites sordidus Germar) outbreaks in the mid-1980s leading to widespread plantation failure in commercial growing areas of Masaka and Rakai districts (Sengooba, 1986; Sebasigari and Stover, 1988); (2) crop decline in the central region; and (3) yield declines already in evidence in parts of the southwest (e.g. Masaka and Ntungamo districts). Therefore, a multi-disciplinary study was conducted at 9 central and 6 southwestern sites to document shifts in highland banana production and to elucidate the driving forces behind these shifts. Field observations were undertaken at central sites to verify farmer perceptions of production decline.

The results of this study, emphasising the dynamics of geographic shifts of highland cooking banana in Uganda, are presented in Gold et al. (1999). This monograph presents regional summaries and detailed data tables from each of the 15 sites.

A multi-disciplinary Rapid Rural Appraisal was conducted at 9 sites in central Uganda which have undergone sharp banana production decline and at 6 sites of banana expansion in southwestern Uganda (Fig. 1). Activities included key informant and group interviews, transect drives and on-farm verification of production constraints. Villages were visited between April and July 1995. Survey methods are presented in detail in Gold et al. (1999).

Site selection. Sites representing banana decline were selected from Iganga, Kamuli, Mukono and Luwero districts using a grid map produced for diagnostic surveys of banana-based cropping systems (Jagtap, 1993; Gold et al., 1994a). Sites of crop expansion were selected at random in Mbarara and Bushenyi districts from traditional pasture or finger millet (Eleusine coracana L. Gaertn) production areas. Four sites had good access to main roads and markets, while 2 sites were remote and inaccessible during the rainy season.

Study hypotheses. It was hypothesised that the causes of banana decline in central Uganda included socio-economic factors (labour, market availability and government policies), pests (weevils and nematodes) and declining soil fertility. Banana expansion in the southwest was hypothesised to be market driven.

Interviews. Key informant interviews on crop production trends and dynamics were conducted with District Agricultural Officers, local parish council members and village elders. Group interviews, using an informal check list, consisted of 25 to 60 respondents representing a cross section of farmers in the village.

Farmers allocated beans to indicate relative importance of the different crops (i.e., land area and home consumption patterns for food crops; land area and income for cash crops). Farmers also listed and ranked causes of banana decline or expansion in their respective regions. Time lines were drawn to show the timing of shifts in banana importance from the 1950s through the 1990s using the following categories as guidelines:

Very low. Banana primarily grown in kitchen gardens or small plots with less than 25 mats.

Low. Banana grown in small stands with few farmers having more than 100 mats, with low productivity and management attention given low priority.

Medium. Banana important but not dominant crop in site with 25 to 50% of farmers growing stands of more than 100 mats. Management attention variable.

High. Banana one of most important crops in village. A majority of farmers had stands of more than 100 mats, and management attention given high priority.

Very high. Banana the dominant crop in village with most farmers having stands of more than 100 mats, and management attention given high priority.

On-farm verification. In each site of banana production decline, 6 farms were randomly selected for case studies. Each farm supported a banana stand more than 2 years old and with more than 50 mats. Most banana stands showed signs of declining production.

Data were collected on socio-economic factors, weevil and nematode damage, and foliar nutrients levels. Weevil damage scores were estimates from two corm cross sections and represent the percentage of corm consumed by weevils. Root necrosis indices represent the percentage of necrotic root material in 5 randomly selected roots from flowered plants. Sampling methods are detailed in Gold et al. (1994, 1999).

Internal weevil damage scored higher than 5% probably reflected meaningful yield reduction, while damage exceeding 10% presented a serious problem. Nematode root necrosis index values greater than 5% are also likely to reflect substantial yield reductions. Densities of 1000-1500 Radopholus similis, 5000 Helicotylenchus and 10000 Pratylenchus goodeyi per 100 g of roots are the best estimates for population thresholds causing significant yield loss (P.R. Speijer pers. comm.).

Figure 1 Map of villge sites for study on geographics shifts in highland cooking banana in Udanda

This study documented the decline of highland cooking banana in central Uganda. Cooking banana has been present in this region for centuries and had been the dominant staple food. For many decades, cooking banana had been a sustainable crop with long plantation life and high stable yields. By the 1990s, bunches weighing more than 10 kg were uncommon and stands often required replanting after only 2-5 years. As a result, farmers in central Uganda have been unable to take advantage of the growing market for cooking banana in urban centres.

Cooking banana production in selected sites in central Uganda fell from 18% of total food crop and 7% of total cash crop production in the 1970s to 4% and 2%, respectively, in the 1990s. Farmers identified reduced labour availability and management, increasing pest pressure and declining soil nutrient status as the major causes of decline. On-farm verification confirmed farmers' observations. Weevil levels were the highest yet found in Uganda, while foliar samples indicated nutrient deficiencies in Mg, N, and K. Soil nutrient deficiencies, however, appear to be a direct outcome of reduced management rather than "soil exhaustion" as postulated by farmers.

A likely cycle of decline was that as labour became more scarce and regional by-laws more lax, management standards declined. Weevil control practices (e.g. sanitation) and soil conservation were abandoned leading to the rise of secondary pest and soil constraints. As banana productivity declined, farmers became discouraged and began turning their resources to other more productive crops. Greater penetration of the cash economy also encouraged farmers to divert their attention to annual crops. In addition, banana growers must now confront new and serious pest and disease problems (i.e., Radopholus similis and black sigatoka) which have arrived in the last 20 years. At present, farmers are pessimistic about the ability to restore banana production in this zone.

Southwestern Uganda has largely presented a mirror image of what has transpired in the central zone. In this area, the traditional staple had been finger millet. By the 1940s banana had been introduced into a few sites only. However, returning migrant labour brought back cooking banana culture from the central zone in the following decades and the crop quickly began to gain favour because it offered better food security than finger millet. Farmers were able to produce large bunches, which attracted traders supplying the growing urban markets in Kampala and Jinja. This further encouraged farmers to expand the land given over to banana and to intensify management practices. In general, management standards in commercial banana growing areas, such as Mbarara and Bushenyi, are high.

Farmers throughout the commercial growing areas of south-western Uganda are optimistic about the future of banana production. Nevertheless, early stages of yield decline are already in evidence in parts of southwestern Uganda (e.g. Masaka district). This may be caused, in part, by the removal of nutrients through marketed bunches, without replenishment through soil amendments.

In summary, cooking banana is a favoured crop in southwestern Uganda because of its high market value and its role as an important staple food. Banana benefitted from government control of the coffee market, suppressing farm-gate prices which discouraged coffee production. At the same time, an expanding urban market made cooking banana became an attractive alternative as a commercial crop. As a result, banana received a high level of management which was reflected in good to excellent yields. However, liberalisation of the coffee market could result in a shift back to coffee and decreased emphasis on banana.

Part of the region benefits from volcanic soils and/or elevation above upper threshold limits for banana weevil, Radopholus similis and black sigatoka pathogen. Sustained high banana yields are likely in these areas, even at lower levels of management. Elsewhere, however, reduced management attention may result in banana decline with accelerated loss of soil nutrients and increasing pest and disease pressure. It will be the challenge of the Uganda National Banana Research Programme to monitor banana production dynamics in southwestern Uganda and preclude the process of decline which led to banana disappearance in central Uganda.

La banane verte (d'altitude) a cuire (Musa spp. Type AAA-EA) est l'aliment de base le plus important dans le sud de l'Ouganda et partout dans la region des grands lacs d'Afrique orientale. Cette culture est une composante-cle a la fois pour la securite alimentaire et la durabilite de l'agriculture dans cette region. Les parcelles de bananiers bien entretenues peuvent souvent subsister pendant trente annees ou plus, meme en condition de faible apport en intrants. De meme, une periode de recolte prolongee assure des aliments et des revenus pendant toute l'annee. Les bananes d'altitude contribuent a la reduction de l'erosion du sol sur les pentes abruptes et constituent des sources principales de paillis en vue du maintien et de l'amyelioration de la fertilite du sol (INIBAP, 1986). Les systemes de production a dominante banane ont fourni l'option la plus durable dans les conditions ecologiques d'altitude moyenne et elevee en Afrique orientale. Dans cette region, l'Ouganda est le premier pays producteur et consommateur de banane.

Entre 1970 et 1990, l'Ouganda a connu une baisse de la production de banane a cuire dans les zones traditionnelles de culture dans la region centrale du pays (Iganga, Jinja Kamuli, Luwero, Mpigi, Mukono. Pendant ce temps, la population non asiatique de Kampala a enregistre une croissance de l'ordre de 750% entre 1960 et 1990. L'augmentation consecutive de la demande en banane a cuire sur le marche urbain a favorise l'expansion de la culture dans les zones encore productives au niveau du pays). Par consequent, la production a ete concentree dans la region du sud-ouest du pays. Actuellement, 63% des terres utilisees pour la production de la banane a cuire se trouvent dans cinq districts du sud-ouest de l'Ouganda : Mbarara (20%), Bushenyi (16%), Masaka (11%), Mubende (8%) et Rakai (8%) (Ministere de l'agriculture de l'Ouganda, 1992).

Les apprehensions relatives a la durabilite de la banane dans les zones de production du moment ont ete mises en exegue par trois facteurs: (1) infestations severes du charaneon du bananier (Cosmopolites sordidus Germar) au milieu des annees 80 ayant cause des pertes de recolte a grande echelle dans les plantations situees dans les zones de production commerciale au niveau des districts de Masaka et de Rakai (Sengooba, 1986; Sebasigari et Stover, 1988); (2) baisse de la production dans la region centrale; et (3) baisse des rendements deja signalee dans certaines zones de la region du sud-ouest (par exemple, les districts de Masaka et de Ntungamo). Par consequent, une etude pluridisciplinaire a ete menee dans 9 et 6 sites respectivement dans les regions du centre et du sud-ouest afin de documenter les changements en matiere de production de banane d'altitude et de mieux comprendre les forces motrices de ces changements. Des observations au champ ont ete realisees dans les sites du centre afin de verifier les perceptions des agriculteurs en ce qui concerne la baisse de la production.

Les resultats de cette etude qui mettent l'accent sur la dynamique des changements geographiques relatifs a la banane (d'altitude) a cuire en Ouganda, sont presentes dans Gold et al. 1999. La presente monographie contient des resumes regionaux et des tableaux de donnees detaillees en provenance de chacun des 15 sites.

Une enquete pluridisciplinaire, selon la methode acceleree de recherche participative (MARP), a ete menee dans 9 sites du centre de l'Ouganda qui ont subi une baisse nette de la production de banane et dans 6 sites d'expansion de la banane dans le sud-ouest de l'Ouganda. Les activites ont porte, entre autres, sur les entretiens avec les enquete-cle, les entretiens de groupe, les eudes transversales et la verification des contraintes a la production en milieu reel. Les villages ont ete visite entre avril et juillet 1995. Les methodes d'enquate sont presentees de maniere detaillee dans Gold et al. (1999).

Choix des sites : Les sites representant la baisse de la production de banane ont ete choisis dans les districts d'Iganga, de Kamuli, de Mukono et de Luwero en utilisant une carte quadrillee etablie pour les enquetes exploratoires sur les systemes a dominante banane (Jagtap, 1993; Gold et al., 1994a). Les sites d'expansion de la culture ont ete choisis de maniere aleatoire dans les districts de Mbarara et de Bushenyi au niveau des paturages traditionnels ou des zones de production de mil. Quatre sites avaient un bon acces aux routes et aux marches principaux, tandis que 2 sites etaient eloignes et inaccessibles pendant la saison des pluies.

Hypothese de l'etude: L'hypothese suivante avait ete avancee: les causes de la baisse de la production de banane dans la region centrale de l'Ouganda comprenaient des facteurs socio-economiques (main-d'oeuvre, disponibilite de marches et les politiques gouvernementales), les ravageurs (charaneons et nematodes) et la baisse de la fertilite du sol. L'hypothese de l'expansion de la culture de la banane quant a elle, etait censee etre induite par le marche.

Entretiens: Des entretiens ont ete effectues aupres des enquetes-cles et ont porte sur les tendances et la dynamique de la production, en collaboration avec les agents agricoles du district, les membres du conseil municipal et les anciens du villages. Les entretiens de groupe ont ete menes, a l'aide d'une liste de controle informelle, aupres de 25 a 60 enquetes representant un echantillon d'agriculteurs du village.

Les agriculteurs ont utilise des graines de haricot pour indiquer l'importance relative des differentes cultures (c'est-a-dire, les superficies et les modes de consommation des cultures vivrieres au niveau du menage; les superficies et les revenus generes par les cultures de rente). Les agriculteurs ont egalement dresse et hierarchise la liste des causes de la baisse de la production de banane ou de son expansion dans leurs regions respectives. Des lignes temporelles ont ete tracees pour indiquer la periode du changement d'importance de la banane, de 1950 a 1990, selon les categories suivantes comme reperes:

Tres faible-Banane essentiellement cultivee dans les jardins potagers ou dans de petites parcelles avec moins de 25 pieds.

Faible - Banane cultivee avec de petits pieds et peu d'agriculteurs ayant plus de 100 pieds; tres peu d'attention a l'entretien et a la faible productivite.

Moyen - Banane en tant que culture - importante mais non dominante dans le site ou 25 a 50% des - agriculteurs cultivent des parcelles de plus de 100 pieds. Attention variable en matiere d'entretien.

Eleve - Banane en tant qu'une des cultures les plus importantes du village. Une majorite d'agriculteurs avaient plus de 100 pieds et priorite elevee en matiere d'entretien

Tres eleve - Banane en tant que culture - dominante du village, avec la plupart des agriculteurs ayant plus de 100 pieds et priorite elevee en matiere d'entretien.

Verification en milieu reel: Dans chaque site de baisse de la production de banane, 9 plantations ont ete choisies de maniere aleatoire pour les etudes de cas. Chaque plantation avait une parcelle de bananiers de plus de 2 ans et avec plus de 50 pieds. La plupart des parcelles ont presente des signes de baisse de production.

Les donnees ont ete collectees sur les aspects suivants: facteurs socio-economiques, degats causes par les charaneons et les nematodes et les niveaux d'elements nutritifs foliaires. L'evaluation des degets causes par les charaneons a ete effectuee a partir de deux echantillons de bulbes qui representaient le pourcentage de bulbes consomme par les charaneons. Les indices de necrose racinaire representaient le pourcentage de materiel racinaire necrotique choisi de maniere aleatoire sur 5 racines au niveau de pieds ayant atteint la floraison. Les methodes d'echantillonnage sont detaillees dans Gold et al. (1994, 1999).

Les degats internes causes par les charaneons ayant atteint un taux de plus de 5% ont probablement cause des reductions de rendement significatives, tandis que les taux de plus de 10% indiquaient un probleme serieux. Les valeurs d'indice de necrose racinaire causee par les nematodes superieures a 5% refletent probablement des baisses de rendement substantielles. Les densites de l'ordre de 1000 a 1500 pour Radopholus similis, 5000 pour Helicotylenchus et 10000 pour Pratylenchus goodeyi par 100 g de racines ont constitue les meilleures estimations des seuils de population causant des pertes de rendement significatives (Speijer. Comm pers).

Cette etude a permis de documenter la baisse de la production de la banane (d'altitude) a cuire dans la region centrale de l'Ouganda. La banane a cuire a ete presente dans cette region pendant des siecles et y constituait l'aliment de base. Pendant plusieurs decennies, la banane a cuire avait ete une culture durable dotee d'une longevite en plantation et de rendements eleves et stables. Dans les annees 90, il etait rare de trouver des regimes de plus de 10 kg et les parcelles necessitaient souvent un repiquage apres seulement 2 a 5 annees. Aussi, les agriculteurs dans cette region centrale de l'Ouganda n'ont pas pu tirer profit de la croissance du marche de la banane a cuire dans les centres urbains.

Dans certains sites du centre de l'Ouganda, la production de banane a cuire a connu une baisse de 18% de la production totale de cultures vivrieres et de 7% de la production totale de cultures de rentes dans les annees 70 a 4% et 2% respectivement dans les annees 90. Les agriculteurs ont identifie la reduction de l'entretien et de la disponibilite de la main-d'oeuvre, la pression croissante des ravageurs ainsi que l'appauvrissement des sols en elements nutritifs comme etant les causes majeures de la baisse. La verification en milieu reel a confirme les observations des agriculteurs. L'incidence des charaneons a atteint des niveaux sans precedent en Ouganda, tandis que les echantillons de feuilles ont presente des carences en Mg, N et K. Toutefois, les carences en elements nutritifs semblent etre une consequence directe de la reduction de l'entretien plutot qu'un "puisement" des sols tel que postule par les agriculteurs.

Le cycle de baisse pourrait etre de au fait qu'au fur et a mesure que la main-d'oeuvre devenait rare et les reglements internes moins stricts, les niveaux d'entretien ont baisse. Les pratiques de lutte contre les charaneons (par exemple, les mesures sanitaires) et de conservation des sols ont ete abandonnees, entrainant l'incidence de ravageurs et de contraintes edaphiques secondaires. Au fur et a mesure que la productivite de la banane a baisse, les agriculteurs se sont decourages et ont commence a consacrer leurs ressources a d'autres cultures plus productives. La percee de l'economie de rente a egalement encourage les agriculteurs a porter leur attention sur les cultures annuelle. En outre, les producteurs de banane doivent faire face, a present, a de nouveaux et serieux problemes de ravageurs et de maladies (Radopholus similis et la cercosporiose) qui se sont poses au cours des 20 dernieres annees. Actuellement, les agriculteurs sont pessimistes quant a la capacite de regeneer la production de banane dans cette zone.

Le sud-ouest de l'Ouganda a connu une situation moins grave que celle de la region centrale. Dans cette zone, l'aliment traditionnel de base etait le mil. Dans les annees 40, la banane a ete introduite dans quelques sites. Les travailleurs saisonniers ont rapporte de la banane a cuire de la region centrale au cours des decennies suivantes et la culture a ete rapidement adoptee parce qu'elle offrait une meilleure securite alimentaire que le mil. Les agriculteurs ont pu produire de grands regimes qui ont attire les commereants fournissant les marches urbains en croissance a Kampala et a Jinja. Cela a davantage encourage les agriculteurs a etendre les supercifies consacrees a la culture de la banane et a intensifier leurs pratiques de gestion. En general les niveaux de gestion des exploitations agricoles dans les zones de production commerciale de bananes telles que Mbarara et Bushenyi, etaient eleves.

Partout dans les zones de production commerciale de bananes dans le sud-ouest de l'Ouganda, les agriculteurs sont optimistes quant a l'avenir de cette production. Cependant, les premieres etapes de la baisse de rendement ont deja ete observees dans des parties du sud-ouest (district de Masaka). Cela peut etre de, en partie, la suppression des elements nutritifs suite a la commercialisation des regimes, en l'absence d'une regeneration grace aux amendements du sol.

En resume, la banane a cuire est une culture de predilection dans le sud-ouest de l'Ouganda a cause de sa forte valeur marchande et de son role en tant qu'aliment de base important. La banane a beneficie du controle gouvernemental du marche du cafe qui a supprime les prix au producteur, ce qui a decourage la production du cafe. En meme temps, le marche urbain en expansion a fait de la banane a cuire une culture commerciale de substitution attrayante. Par consequent, la banane a beneficiee d'un niveau de gestion eleve qui s'est traduit par des rendements variant de bons a excellents. Toutefois, la liberalisation du marche du cafe pourrait entrainer un retour de cette culture au detriment de la banane.

Une partie de la region beneficie de sols volcaniques et/ou eleves au dessus des seuils d'incidence du charaneon du bananier, Radopholus similis et de la cercosporiose. Dans ces zones, la durabilite des rendements eleves est probable, meme en condition de faible entretien. Toutefois, ailleurs, la reduction de l'attention en matiere de gestion pourrait entrainer une baisse de la production ainsi qu'une perte acceleree des elements nutritifs du sol et l'augmentation de la pression des maladies et des ravageurs. Il reviendra au Programme national de recherche sur la banane de l'Ouganda de relever le defi suivant: assurer le suivi de la dynamique de la production de banane dans la region sud-ouest et prevenir le processus de baisse qui a entraine la disparition de la banane dans la region centrale de l'Ouganda.

A. Central Uganda. To document banana production decline in central Uganda, 9 villages were selected at random from Iganga, Kamuli, Mukono and Luwero districts using a grid map produced for diagnostic surveys of banana-based cropping systems (Jagtap, 1993; Gold et al., 1994). Stratification in the area of interest was based on rainfall and market access (Table 1). At least 3 days were spent in each village.

Farming systems were characterised by high food and cash crop diversity. Study villages averaged 12 food and 11 cash crops. Six of the 9 villages had ranked banana as their leading staple in the 1970s (Table 2). In two other villages, Namugongo and Nakawomero, banana had been an important food crop prior to the 1970s. Cooking banana's importance as a staple food declined in all 9 study sites between the 1970s and 1990s. By the 1990s, banana ranked no higher than fifth as a food crop. In Busembatya, Namugongo and Nakawomero, banana decline was already in evidence in the 1960s, while elsewhere decline began in the 1970s. In most sites, the process of decline continued into the 1990s. Currently, banana production was rated as medium in 2 sites, low in 2 sites and very low in 5 sites. Beans, cassava, maize and sweet potato all gained in importance in the central region between the 1970s and 1990s (Table 2). Cassava was the principal replacement crop for banana and became the leading staple at 7 sites; however, at Busembatya and Nakowomero, cassava use was short lived owing to the rapid spread of Cassava Mosaic Disease. Maize had become the primary staple at 3 sites, including Busembatya, while sweet potato was the most important food in Nakowomero. During the same period, rice moved from 15th to the third most important food crop in Namugongo. Groundnut remained of moderate importance at most sites where it was used as a sauce to accompany starchy staples.

Cooking banana was the second most important cash crop at Bulyangada and Busagazi during the 1970s but was never an important commercial crop at the other sites (Table 3). Declining yields coincided with the period of urbanization in Uganda, making it impossible for banana producers in this region to compete for the growing urban market. As a result, banana marketing infrastructure remained poor since the low production could not attract traders. Coffee and cotton had been the leading staple crops during the 1970s (Table 3). Governmental control of the coffee market resulted in suppressed farm-gate prices and discouraged production. By the 1990s, coffee's importance had declined by half although it remained the leading staple crop in 4 sites. At the same time, cotton had virtually disappeared due to degeneration of the cotton marketing system. Maize, beans and beer banana all showed substantial increases in importance during this period (Table 3). For example, brewing banana's share rose from 1% of cash crops in the 1970s to 12% in the 1990s.

Table 1. Sites selected in central Uganda for a Rapid Rural Appraisal to document shifts in banana production in Uganda, May, 1995

Table 2. Shifts in importance of primary food crops at selected sites in central Uganda between the 1970s and 1990s

1 = most important

Table 3. Shifts in importance of primary cash crops in selected sites in central Uganda between the 1970s and 1990s

1 = most important

Central Uganda is characterised by increasing land pressure, small farms and resource-poor farmers. At the different sites, farmers presented a list of 4 to 9 factors contributing to banana production decline in their village. The most important of these were increased pest pressure, reduced management attention, declining soil fertility, climatic changes (increased drought and unreliability of rains) and land pressure (Table 4). In Mpoma, off-farm employment or trade resulted in reduced management attention for all crops.

Farmers reported increasing pest pressure (especially banana weevil) as a cause of rapid banana decline. Some farmers attributed increased pest problems to reduced management attention and lack of field sanitation (i.e., weeding and removal of crop residues). Field observations verified farmer perceptions that the weevil was an important constraint: Damage levels on study farms were the highest yet found in Uganda (Table 5) and 3 times the country-wide average (Gold et al., 1994). The data suggest yield loss of 50% or more if these levels of damage are sustained over several crop cycles (cf. Rukazambuga et al., 1998).

Although most farmers were unaware of nematodes, Radopholus similis and Helicotylenchus multicinctus also appear to be important constraints (Table 5). Their attack resulted in high root necrosis indices and reduced root numbers, suggesting potential yield loss from nematodes of 20-40% (P.R. Speijer pers. comm.).

Reduced management attention was attributed to 1) abandonment of government by-laws; 2) decreased labour availability; and 3) shifted attention to other crops or activities. Government by-laws which once mandated mulching, field sanitation, rotation and fallowing went unenforced during the political turmoil of the 1970s and 1980s. Farmers in the central region also were dependent upon external labour (readily available from Rwanda and southwestern Uganda) for banana stand management. Declining availability (and increasing costs) of labour in the 1970s resulted in lower management attention, reduced yields, and shifted resource allocation to more productive crops.

Banana was reported as the crop most affected by soil fertility decline, while maize, cassava, beans and sweet potatoes continued to perform well. Despite the low market incentive to banana farmers, cooking banana remained the preferred food in the region. As a result, banana was restricted to the farmers's best land in spite of its relative low management priority.

Table 4. Farmer perceptions of factors contributing to banana production decline in central Uganda

1 = most important

Table 5. Weevil and nematode levels on selected farms in central Uganda

Key
XI: Percent damage to central cylinder
XO: Percent damage to outer cortex
XT: Percent damage to corm
RI: Root necrosis index
RS: Numbers of Radopholus similis
HM: Numbers of Helicotylenchus multicinctus

Clonal diversity of highland cooking banana (matooke) remained high (average 13/site) in central Uganda (Table 6), in spite of the reduced importance of the crop. Virtually all of the clones being grown in the 1970s were still being grown, although 32% of the clones were reported to be "disappearing". Highland brewing banana (mbidde) was of minor importance with only two sites continuing to maintain clones (1/site).

Angeles et al. (1993) summarised the critical nutrient values for dessert bananas reported in 26 studies and used these data to determine Diagnosis and Recommendation Integrated System (DRIS) norms (Table 7a). In this study, DRIS analysis of foliar data suggests that magnesium was deficient at all sites and was the key constraint in the region, while banana stands were also deficient in potassium (6 sites) and in nitrogen (5 sites) (Table 7b). This raises the questions as to whether these nutrient deficiencies are of recent origin and, in turn, have played an important role in cooking banana decline.

Historical data on banana foliar nutrient levels in central Uganda are limited. Parish (1969) reported yields of 20 tons ha-1 as common in central Uganda. He also described nitrogen and phosphorus levels as "marginally low", while potassium was "very low" and a major constraint to yield (Table 7c). Thus, foliar nutrient deficiencies appear to have been present in the central region prior to the onset of banana decline.

Nevertheless, the results suggest that farmers were correct in their assessment that diminishing soil fertility (in particular, magnesium) has contributed to lower banana yields. However, it is unlikely that the rate of "soil exhaustion", as described by farmers, could explain 50-75% declines in banana yield over a 20-30 year period. Thus, it appears more likely that problems associated with reduced management (including pests), rather than soil exhaustion from overcropping, were the primary causes of banana decline in central Uganda.

B. Southwestern Uganda. Sites representing areas of highland banana production expansion were selected at random in Mbarara and Bushenyi districts from traditional pasture or finger millet areas. Four sites had good access to main roads and markets, while 2 sites were remote and inaccessible during the rainy season.

Farming systems were characterised by high food and cash crop diversity with study sites averaging 10 food and 9 cash crops. Traditionally, finger millet had been the leading staple crop for much of southwestern Uganda, while highland banana had been relegated to kitchen gardens. Considerable land was also given over to pasture for raising livestock.

Banana first gained importance in Biharwe in the 1940s, in Katooma and Kabuyanda in the 1950s, in Kashaka in the 1960s and in Buhweju and Kagarama in the 1970s. By the 1970s, banana was the first or second most important staple crop in 4 of the 6 southwestern sites (Table 9). Only in the 2 remote sites, Buhweju and Kabuyanda, was banana relatively unimportant at that time. Currently, banana is the leading staple crop in all study sites. Overall, banana's share as a food crop rose from 20% in the 1970s to 26% in the 1990s (Table 9).

Expansion of banana was primarily at the expense of finger millet, which declined in importance at survey sites from 19% in the 1970s to 11% in the 1990s (Table 9). During the same period, there was virtually no change in beans and cassava at any site, while maize production declined at two sites and increased at two other sites.

Marketing of cooking banana increased dramatically throughout the region over the last 20 years. During the 1970s, banana was an important cash crop in only 1 site (Katooma) and not marketed at all in 3 villages (Table 10). In contrast, by the 1990s, banana was the leading cash crop at 4 sites. Overall, banana share as a cash crop rose from 6% to 27%.

TABLE 6. Highland banana clonal diversity and status at selected study sites in central Uganda

Table 7. Foliar nutrient levels on study farms in central Uganda sites undergoing highland banana production decline

1 Average = mean of 26 published studies; source: Angeles et al. (1993)
2 Source: Parish (1969)
(Adopted from Gold et al. (1999)

Table 8. Sites selected in southwestern Uganda for a Rapid Rural Appraisal to document shifts in banana production in Uganda, May, 1995

TABLE 9. Shifts in importance of primary food crops at 6 selected sites in southwestern Uganda between the 1970's and 1990's as evaluated in a Rapid Rural Appraisal, May - June 1995

1 = most important

Banana was of only moderate importance as a cash crop at the two remote sites, Buhweju and Kabuyanda (Table 10). Buhweju had virtually no market access in the 1970s; with somewhat better roads, it has recently developed a speciality market for peas. Kabuyanda has access to an alternative market in Kabale for beans and maize. In both sites, infrequent and sporadic visits of traders make cooking banana a moderate risk crop. Therefore, farmers rely more on beer banana for gin (waragi) production because of its longer shelf life. Beer banana was the second and third most important cash crop in Buhweju and Kabuyanda, respectively.

Increases in banana production in southwestern Uganda coincided with banana decline in the central region, with increased production underway in 5 sites during the 1970s. By the 1990s, production was described as high (1 site) or very high (5 sites). Thus, banana expansion in the region was primarily as a cash crop. In some cases, banana out-competed and replaced other cash crops. At other sites, expanding markets encouraged an increase in crop acreage without substitution of banana for other cash crops. Prior to the 1970s, coffee had been the major cash crop of the region. By the 1990s, coffee had largely disappeared (Table 10) because of unfavourable marketing policies. Shifts in other cash crops, including beans, groundnut, maize, finger millet, onions, peas, sorghum, tomatoes varied by site, increasing in importance in some sites, while decreasing in others.

Southwestern Uganda is characterised by a mixture of commercial growers and resource-poor farmers. Farmers listed increased market demand, food security and improved management as the primary factors contributing to increased banana production in the region (Table 11). All study villages served as sources of migrant labour to central Uganda in the 1960s. These workers returned with "banana culture" which eventually gained favour over finger millet due to ease of production and preparation, stability of yield and year-round harvest (food security). High yields coincided with the emergence of urban markets; traders were attracted to southwestern Uganda where production of large marketable bunches was common. Marketability of the crop resulted in expansion of banana acreage and increased management attention to sustain high yields. Moreover, in contrast to central Uganda, farmers in the southwest employ solely their own labour for banana management and have developed a relatively high standard of agronomic practices.

Diversity of cooking banana clones (mean of 11 per site) was similar to that in central Uganda (Table 12). Few new clones had been introduced since the 1970s and about 2 clones per site were being phased out. All brewing banana (mbidde) clones were reported as being phased out, including at Buhweju and Kabuyanda where banana beer remained an important cash commodity.

Table 10. Shifts in importance of primary cash crops at 6 selected sites in southwestern Uganda between the 1970's and 1990's

1 = most important

Table 11. Farmers perceptions of factors contributing to increasing banana production in southwestern Uganda

Table 12. Clonal diversity and status in southwestern Uganda

Copyright © 1999 The African Crop Science Society

The following images related to this document are available:

Line drawing images