Edited by: Giovanni Bubici, National Research Council (CNR), Italy
Reviewed by: Renata Retkute, University of Cambridge, United Kingdom
Naitong Yu, Chinese Academy of Tropical Agricultural Sciences, China
A/Prof. Thomas, University of Queensland, Australia
*Correspondence: Walter Ocimati,
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Banana bunchy top disease (BBTD) caused by banana bunchy top virus (BBTV) poses a significant threat to Uganda’s food and income security.
To map BBTD spread and inform its management, a delimiting survey was undertaken covering the high-risk zones bordering the Democratic Republic of Congo (DR Congo) in the west, Rwanda and Tanzania in the south, and South Sudan in the north. BBTD is endemic in the DR Congo and present in Tanzania and Rwanda. The survey and environmental data were then used to map the vulnerability of Uganda’s banana landscapes.
BBTD was only confirmed on 9% of sampled farms in north- and midwestern Uganda, with yield losses of 75% to 100%. Farmers observed BBTD over a 0.5–4.4-year period, suggesting a delayed detection. Suckers were the predominate planting materials used, increasing the risk of disease spread. Landscape suitability for BBTD was influenced by precipitation of the driest month, banana presence in 2016, land surface temperature difference (LSTD), the interaction between wind speed and LSTD, isothermality, wind speed, and the normalized difference vegetation index. These variables affect either or both the virus and aphid vector populations. Altitude did not influence the model, possibly due to disease introduction at mid to high altitudes through infected planting materials. The low-lying zones (around River Nile and Lakes Albert, Edward, and Victoria) are highly vulnerable. BBTD risk was low in northeastern Uganda with low banana production. The prediction map shows some suitable landscapes in the southwest that can expose this major banana production zone to BBTD, necessitating proactive measures.
香京julia种子在线播放
Banana bunchy top disease (BBTD) is a viral disease of bananas and plantains (
Severe BBTV infection is characterized by stunted growth with distorted leaves having a bunchy and erect appearance in young plants, while bunches of mature plants fail to fill, as such fruits remain inedible (
In Uganda, bananas are a vital staple food for a significant proportion of the population, providing essential nutrients and contributing to food security. Uganda, with a daily per capita consumption of approximately 0.7 kg/person, ranks highest in the consumption of bananas in the world (
The risk of BBTD spread to other banana production zones within Uganda remains high due to several factors including the lack of awareness on BBTD, its symptoms, epidemiology, and control; heavy reliance on suckers sourced from farmers’ own or neighboring farms for planting materials; lack of proper quarantine measures; and limited resources for implementing appropriate control measures. BBTV is known to spread through infected planting materials and via the aphid vector
In Uganda, management of the disease in currently infected landscapes and prevention of further spread into new production zones require a proper understanding of the vulnerability of the different banana production landscapes to disease introduction and/or establishment. Thus, mapping of the vulnerability of banana production landscapes to BBTD is an essential step toward developing targeted management strategies and implementing preventive measures. By identifying areas at higher risk of disease introduction and/or establishment, stakeholders can prioritize resources and interventions, thereby reducing the overall impact of BBTD on banana production systems. Moreover, mapping vulnerability will aid in decision-making processes, such as implementing disease surveillance programs, designing strategies for disease containment and eradication, and prioritization of limited resources.
This study highlights the findings of initial diagnostic studies conducted on sampled farms in the affected regions and provides a prediction of the risk/vulnerabilities of the different banana production landscapes in Uganda to the establishment and local spread of BBTD. The study utilized a combination of field diagnostic surveys and spatial analysis, to assess past distribution patterns and factors influencing the vulnerability of the banana ecosystem in Uganda to BBTD.
This study builds on the first report of BBTD in the West Nile region of Uganda (
Diagnostic surveys were conducted in 46 districts of Uganda located in the border zones with South Sudan, the DR Congo, Rwanda, and northwestern Tanzania (
Uganda map showing the surveyed regions and districts.
A combination of stratified purposive sampling and respondent-driven snowball sampling techniques was employed in this study. A stratified purposive sampling was attained by selecting, with guidance from the district and subcounty agricultural extension staff, subcounties, and parishes with high banana production and/or bordering BBTD hot spots. Farms in these subcounties were randomly visited and sampled through the snowball sampling technique. Snowball sampling is a recruitment technique through which research participants assist in identifying or enlisting other potential subjects (
Banana plants with typical banana bunchy top disease symptoms.
The interviewers examined plants within the fields to confirm the presence or absence of the disease. Where the disease was observed or reported in the neighborhood by farmers, through the snowball approach, these suspected farms were also visited. Samples were collected from plants with characteristic BBTD symptom(s) for laboratory analysis to confirm the presence of the BBTV. BBTV was confirmed with the help of a PCR using primers that amplify a 240-bp conserved domain in the BBTV DNA-mRep segment (
The datasets were cleaned and preprocessed in Microsoft Excel (version 2302, build 16130.20332), and ANOVA determined using RStudio software (version 2023.3.0.386;
For BBTD risk mapping, the entire country of Uganda, covering an area of approximately 241,038 km2, was used. Uganda borders South Sudan in the north, Kenya in the east, Tanzania and Rwanda in the south, and the DR Congo in the west.
Datasets used for BBTD risk mapping consisted of field survey results and maps/databases of key covariates known to influence BBTD spread. Data from three field surveys conducted between 2020 and 2022 that determined the presence and/or absence of BBTD along high-risk banana-producing areas in Ugandan districts bordering the DR Congo, South Sudan, Rwanda, and Tanzania were used. The first two surveys concentrated on the northwestern region of Uganda that already had reported BBTD cases, while the third survey stretched from the border regions with South Sudan in the north, DR Congo in the west, and Rwanda and Tanzania in the south. The three surveys generated a total of 630 survey data points. All plants reported to have BBTD were checked for BBTV using PCR, and only farms with plants confirmed as infected were retained as positive. The field survey datasets were subjected to cleaning and preprocessing using Microsoft Excel (version 2302, build 16130.20332), while shape files were processed using QGIS (version 3.30.0, Hertogenbosch).
The covariate datasets were downloaded as raster files (“.tiff”) from publicly available databases (
Initial 33 covariates used in the study.
Category/covariate (code) | Description | Source | Resolution |
---|---|---|---|
Topo-Graphic | |||
Elevation (Elev.0) | Altitude [masl] |
|
30 arcsec (~1 km2) |
Climatic | |||
• Annual mean temperature [°C] (TempA.1) |
1970–2000 |
|
30 arcsec |
• Annual mean precipitation (PrecA.12) |
1970–2000 |
|
30 arcsec |
Solar radiation (srad.20) | kJ m−1 day−1; annual mean from January to December |
|
30 arcsec |
Wind speed (wind.21) | ms−1; annual mean from January to December |
|
30 arcsec |
Water vapor (vapr.22) | kPa; annual mean from January to December |
|
30 arcsec |
Recent annual precipitation (Newprec.23) | Mean between 2019 and 2021 |
|
30 arcsec |
Recent minimum annual temperature (Newtmin.24) | Mean between 2019 and 2021 |
|
30 arcsec |
Recent maximum annual temperature (Newtmax.25) | Mean between 2019 and 2021 |
|
30 arcsec |
Land surface temperature difference (LSTD.26) | Annual mean from January to December |
|
0.1 ○ |
Vegetation | |||
Normalized difference vegetation index (NDVI.27) | Mean of months between Jan 2022 and Jan 2023 (terra/MODIS) |
|
0.1 ○ |
Leaf area index (LAI.28) | Mean of Jan and Feb 2017 |
|
0.1 ○ |
Vegetation [land cover] (Cover.33) | Africa 2016 |
|
30 arcsec |
Banana production | |||
Banana presence in 2016 (ProdB16.29) | Banana presence in 2016 |
|
|
Attitude (Altitude.ug) | This study | ||
Longitude (Longitude) | Derived from SRTM v4.0 elevation data | 30 arcsec | |
Latitude (Latitude) | Derived from SRTM v4.0 elevation data | 30 arcsec |
Covariate processing and data analysis were done in RStudio software (version 2023.3.0.386;
Most of the covariates were imported as objects into RStudio at a resolution of 30 arcsec (~1 km2). Those that were not to this projection were first re-projected to the elevation using the “projectRaster” function of R. Banana presence raster of 2016 (
All variables were masked to the Ugandan polygon shape which is in the WGS 1984 coordinate system and then to the elevation raster to ensure a similar extent, with a bilinear method. The cropped and masked rasters were written as rasters (“writeRaster” function) for future use as Ugandan raster files for plotting covariate maps.
The data points of each covariate at the surveyed location were first extracted by transforming the surveyed datasets into a data frame of a coordinate system (longitude, latitude, and altitude) using the “extract” function of the raster package in R. The Pearson correlation analysis was performed to establish correlations between all the variables. Correlation analysis was performed to select a single covariate to represent those that are highly correlated or collinear (
Regression models use the relationship between the dependent variable and the covariates to predict at the unobserved locations. The covariates from the univariate logistic regression analysis with a probability significance of
Covariate maps for the eight variables used in the multivariate regression analysis.
These covariates were then entered into a multivariate logistic regression model using the glm function with the method of “logit.” Using stepwise regression, on both forward and reverse directions, the covariates that were not significant at 5% (
and
Where
The final derived logistic regression model was stacked or applied to the final covariate maps to obtain the regression prediction map showing the risk of spread/environmental suitability of Ugandan landscapes to the spread and establishment of BBTD. The risk of BBTD spread and/or likely distribution in the logistic regression model was therefore expressed by the binary response variable of “1”s and “0”s representing BBTD presence and absence, respectively, regressed to other covariates.
A total of 630 banana farms were surveyed across four study regions (i.e., northwestern, midwestern, southwestern, and central regions) of Uganda. The mean area under banana production across the study sites was 0.4 ha. The sampled farms were located at varying altitudes, from as low as 604 meters above sea level (masl) to as high as 2,044 masl (
Banana field management practices significantly differed (
Respondents using different cropping systems
Banana bunchy top disease was observed on approximately 9% of visited farms in the northwestern and midwestern regions. No BBTD was confirmed in the other study regions (
Map showing the presence (red dots) and absence (green dots) of BBTD in the surveyed districts in Uganda.
Farmers responded differently to banana diseases across the study regions. Cutting single diseased stems was the commonest response to diseases on banana farms and was applied on 57%, 91%, and 100% of the farms in the north-, south- and midwestern regions, respectively (
The response of farmers to banana diseases across the surveyed banana landscapes in Uganda.
Pearson correlation coefficients between the covariates were generally high, especially within covariates that measured a common variable, e.g., temperature and precipitation (
Correlation analysis of all 33 covariates and the dependent variable (disease presence/absence) using Pearson’s method. TempDr.2, annual mean diurnal range; isotherm. 100.3, isothermality; TempAr.7, temperature annual range; TempWeq.8, temperature of the wettest quarter; PrecA.12, annual mean precipitation; PrecWm.13, precipitation of the wettest month; PrecDm.14, precipitation of the driest month; PrecV.15, precipitation seasonality; PrecWaq.18, precipitation of the warmest quarter; PrecCq.19, precipitation of the coldest quarter; srad.20, solar radiation; wind.21, wind speed; vapr.22, water vapor; LSTD.26, average land surface temperature; NDVI.27, normalized difference vegetation index; LAI.28, leaf area index; ProdB16.29, banana presence (
A total of eight covariates were entered into a multivariate logistic regression with BBTD presence/absence. Of the eight covariates, two covariates (annual mean diurnal temperature range, precipitation of coldest quarter) were dropped because they did not contribute enough to lowering the deviance of the model. The model was explained by six covariates and one interaction between two covariates (
Estimates, standard errors, and
Covariate | Estimate | SE | Pr(>| |
---|---|---|---|
(Intercept) | −197.8190 | 89.6144 | 0.0273* |
Isothermality (isotherm) | −0.1326 | 0.0685 | 0.0527 |
Precipitation in the driest month | 0.2155 | 0.0527 | 4.35E−05*** |
Wind speed | 88.7381 | 46.3263 | 0.0554 |
Normalized difference vegetation index | −0.0421 | 0.0264 | 0.1108 |
Land surface temperature difference (LSTD) | 1.1111 | 0.4592 | 1.55E−02* |
Banana presence in 2016 ( |
2.5178 | 1.1035 | 0.0225* |
Wind speed: LSTD | −0.4871 | 0.2383 | 0.0409* |
“*,” and “***” denote significance at p ≤ 0.05, and p ≤ 0.001, respectively. Null and residual deviances are, respectively, 323 on 621 degrees of freedom (df) and 245 on 614 df. AIC: 261.
The BBTD risk map (
Likelihood of establishment and spread of BBTD within different landscapes in Uganda.
Bananas are an important food and income crop in Uganda with high production concentrations, respectively, in the southwestern, midwestern, and central regions of Uganda. Banana production in the country has been greatly constrained by Xanthomonas wilt of banana, a bacterial disease. The report of BBTD in the northwestern part of the country (
The banana production systems were dominated by smallholders with a mean farm area of 0.4 ha. Natural capital such as access to land greatly influences the farmers’ decisions to adopt or dis-adopt innovations and may have implications on future efforts to manage BBTD.
The farms in the midwestern and central regions and much of northern Uganda are in low to mid altitudes which are suitable for aphid-vectored BBTV transmission (c.f.
Intercropping though at differing levels of intensity dominated across the studied zones. This could benefit the landscapes through suppression of the BBTV aphid vector. Crop diversification through, e.g., intercrops promote the buildup of a high functionally diverse population of natural enemies, reducing pest densities (
The presence of BBTD over a period of 0.5 to 4.4 years on at least 9% of the farms in the northwestern and mid-southern region suggests that the disease was within the landscapes for a long period of time before detection and official reporting. The fact that some farms had just observed the disease is evidence that the disease is actively spreading within landscapes. This coupled with the observed high yield losses (75% to 91%) highlights the need for urgent action to contain the disease. Symptoms were observed in diverse banana varieties and genotypes including the East African highland banana that are the most important and cultivated group in East Africa. Though the survey did not exhaustively observe all the banana varieties in the country,
Farmers’ responses to diseases including BBTD were documented as a measure for their likelihood to adopt BBTD control measures. Cutting of single disease stems was common across the study sites. This can be attributed to the wide promotion of the practice for managing banana bacterial wilt disease, which is prevalent in the region (
In line with earlier reports (
Temperature affects the aphid vector biology and movement, virus spread and transmission efficiency, and the rate of disease symptom development (
High precipitation, despite improving the suitability of the host to the BBTV aphid vectors, raindrops, and runoff water directly dislodge aphids off the plants, thus reducing their populations on plants (
The negative association of BBTD presence with wind speed could be attributed to the direct dislodgement of BBTV aphid vectors by wind. Previous studies, however, paint a mixed picture of the effect of wind on aphids.
As expected, disease presence was positively correlated to the banana production variable, i.e., banana presence in 2016. The high presence of the host plant creates a suitable environment for the BBTV aphid vector to thrive and thus a higher chance for BBTV spread. Normalized difference vegetation index had a strong negative correlation with BBTD presence. NDVI quantifies the health and density of vegetation and could reflect the role of a healthy and diverse vegetation in creating a suppressive landscape toward the aphid vector. This result, however, contrasts with the findings of
Unexpectedly, a weak negative correlation was observed with altitude (c.f.
Only six covariates and the interaction between wind speed and LSTD had a significant (i.e.,
The LSTD had a strong and positive influence on BBTD presence. Uganda experiences moderate temperatures throughout the year, with monthly minimum air temperatures of 16.6°C to 18.1°C, monthly maximum of 28.0°C to 31.2°C, and monthly average air temperature varying between 22.3°C and 24.3°C. These temperature conditions are within the conducive range for the survival and spread of BBTV and its aphid vector (
Though altitude did not contribute to the final model, it is highly and negatively correlated to the temperature covariates including LSTD (
The BBTD risk map shows swaths of production zones in the northwestern and central Uganda to be suitable for the disease to establish itself. The low-lying production zones with higher temperatures along the River Nile and around Lake Victoria are highly prone to BBTD. Potential BBTD hot spots also exist in the western and southwestern districts that account for the highest banana production in Uganda.
BBTD control has been reported to be extremely challenging once established due to the lack of easy environmentally and economically sound control measures and the omnipresence of the BBTV aphid vector in all banana production landscapes (
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
WOc: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. GO: Data curation, Formal analysis, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing. AT: Data curation, Investigation, Writing – review & editing. JK: Conceptualization, Investigation, Methodology, Project administration, Resources, Supervision, Writing – review & editing. JT: Conceptualization, Investigation, Methodology, Resources, Supervision, Writing – review & editing. RE: Data curation, Investigation, Writing – review & editing. WOk: Data curation, Investigation, Writing – review & editing. GM: Conceptualization, Methodology, Writing – review & editing. DK: Investigation, Writing – review & editing. GB: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Writing – review & editing.
The author(s) declare financial funding was received for the research, authorship, and/or publication of this article. This study was conducted with the financial support from the One CGIAR Initiative on Plant Health and Rapid Response to Protect Food Security and Livelihoods (Plant Health Initiative). We would like to thank all funders who have supported this research through their contributions to the CGIAR Trust Fund:
The authors are grateful for the funding from the One CGIAR Initiative on Plant Health and Rapid Response to Protect Food Security and Livelihoods. The logistical support from the Ministry of Agriculture, Fisheries and Animal Industry (MAAIF) – Uganda and the National Agricultural Research Organization (NARO) of Uganda are also acknowledged. The authors also acknowledge the support from district- and subcounty-level agricultural officers and the banana farmers who willingly participated in this study.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
The Supplementary Material for this article can be found online at:
Altitude at which sampled farms are located