MusaNet members have recently organized and/or been involved in the highlighted projects below.

The Taxonomic Reference Collection (TRC) is a set of accessions representing the main subgroups of cultivated bananas and the two wild species that are the origin of most domesticated bananas, Musa acuminata and Musa balbisiana. The TRC was selected by the Taxonomy Advisory Group (TAG) of MusaNet in 2008 in order to represent as much as possible the entire spectrum of edible banana diversity and to serve as a tool to help resolve some taxonomic issues and improve the usability of the morphological characterization descriptors.

The main objectives of the TRC project are

  • To test the robustness of the standardized descriptors
  • To identify environmental factors that may influence plant morphology (by testing the TRC at field collections across different environments)
  • To promote the fully characterized accessions of the TRC as reference set representing the full Musa diversity

The TRC serves as an important training tool for students, curators, breeders and banana workers as it offers a broad overview of the diversity of bananas.

The collection currently comprises 29 accessions (originally 34) that are conserved and available for distribution at Bioversity’s International Musa Germplasm Transit Centre (ITC) in Belgium. The accessions were planted in 12 partner field collections and are currently being documented and characterized using 1) the minimum set of Musa descriptors (TAG 2016) and 2) the full set of Musa descriptors (IPGRI/CIRAD 1996).

Field collection partners involved

  • BPI (the Philippines)
  • CARBAP (Cameroon)
  • CORBANA (Costa Rica)
  • EMBRAPA (Brazil)
  • FAVRI (Vietnam)
  • NARO Uganda)
  • NRCB (India)
  • IITA-Ibadan (Nigeria)
  • ITFRI (Indonesia)
  • IRAZ (Burundi)
  • SDR-MAP (French Polynesia)
  • USDA (Puerto Rico)


While Musa’s species diversity is beginning to be well defined, the precise ranges of distribution of the different banana groups and subgroups are much less well defined. However, knowledge of this geographical organization is invaluable for many studies. For instance, this knowledge could help:

  • to bring elements to understand and decipher the history of the diffusion of Musa, for instance its diffusion to Africa from SE Asia,
  • to obtain a precise picture of the natural geographical diversity of Musa to foresee the evolution of cultivating areas in connection with expected climate changes and to identify collecting gaps, endangered diversity areas and more globally,
  • to map the distribution of CWRs and landraces in primary and secondary centers of diversity


These analyses rely on the constitution of a geo-located knowledge base of the in-situ repartition of wild and cultivated bananas, with a focus on landraces growing in close to a natural ecological context (in opposition to mass production, with the high and recent introduction of exogenous genotypes). To enrich the database, we need a collection of well-defined banana points across the world. What we mean by “well-defined point” is a place where a banana plant has been recorded and determined (up to the subgroup/subspecies level) with accurate geo-location data (GPS or precise address, a village for example).

Data points are gathered from different sources. Bibliography, including collecting missions’ reports, is analyzed. Online databases, such as Flickr, are monitored to retrieve all relevant observations.  The first results show a very uneven distribution of points according to the regions of the world. To remedy this, the idea of setting up a participatory research project based on a network of experts and amateur volunteers emerged. The aim is to obtain the best possible coverage of banana biodiversity wherever it grows naturally.

For that purpose, a public project has been set up at the iNaturalist website (screenshot above) that may be joined by any volunteer willing to help enrich the database: “Banana natural biodiversity mapping” ( iNaturalist is a citizen science project and online social network of naturalists. Citizen scientists, and biologists built on the concept of mapping and sharing observations of biodiversity across the globe. Observations may be added via the website or from a mobile application. Users of iNaturalist have contributed over eight million observations so far since its founding in 2008, and the project has been called “a standard-bearer for natural history mobile applications.” (Wikipedia). iNaturalist has developed a mobile application (Android & iOS) with a very intuitive user interface which makes it easy to add new observations to the project.

All observers will automatically become part of the project, be able to follow the evolution of data collecting and to engage with other participants and specialists about observations, identifications, planned analyses etc. We provide here some guidelines about the type of observations needed and some basic instructions about the use of the iNaturalist website and application.

In order to create a first core group of specialists capable of carrying out these observations, we turn to MusaNet. The idea is to be able to call on experienced observers, located in the different regions where bananas grow naturally, connoisseurs of their local environment and – of course – passionate about bananas. This first circle could also be a local driving force for other vocations, for example through school and university networks, thus increasing the number of observers. This project is truly a citizens’ science project, one of the objectives of which is to raise public awareness of the importance of knowledge and conservation of genetic resources.

Which plant to record?

Depending on the objective of the study, pictures requirements may vary, but in the end, all recorded banana biodiversity is interesting. For instance, niche modelling relies on large samples of accessions belonging to the same subspecies/subgroup. We will then preferably focus here on common triploid bananas, not rare specimens where we are likely to obtain more occurrences. But if a sufficient number of observations is available, some more unusual specimens may also be recorded for that purpose. On the other hand, observing rare specimens will have great value to have an idea of the distribution of crop wild relatives and landraces. All types of banana are then welcome! The main problem here will be the correct botanical identification of the plant. Hence, taking detailed pictures (see below) is most important.

In any of the above cases, we are looking for plants growing in the closest to natural possible environment. Hence, not on your balcony, not in Iceland, not in a greenhouse, not in an intensive commercial plantation, not in your collection… But maybe in a backyard garden, escaped from cultivation in the wild, on the roadside, in a (very) small farmer plantation…

For instance, recording a Cavendish plant in a large plantation in Ecuador, Costa Rica or Ivory Coast, or under a greenhouse in the Canarias won’t tell us much because plants in this case are irrigated, fertilized, spread… and hence grown in rather artificial environmental conditions. Conversely, knowing that a plantain is grown on the side of a road somewhere in the lowlands of the Littoral region of Cameroon or in the heights of Bujumbura is accurate and useful information. In fact, the most effective way to find these points is certainly to ask authorized people with skills in banana taxonomy. And that’s where we are now!

Picture requirements

No bunch? Then no picture! And no observation… You don’t need to be a banana specialist to add new points to the database. However, we will need to be able to determine the classification of the banana plant from your picture. Without a bunch, it is very difficult, and we would have to discard your observation. The photo to the left is a good example, displaying a bunch of Pisang Rajah. If you are interested on providing more photos, Bioversity International (via MusaNet) has issued guidelines for taking a set of pictures that are useful for botanical determination .

iNaturalist website and application

It is possible to add new observations directly through iNaturalist website ( To do this, it is necessary (but easy) to open an account to save your observations. When adding an observation this way, you will need to indicate as precise as possible GPS coordinates.

Once your observation is added, you have to add it to our project. Just begin to type the name of the project (“Banana…”) under “Projects” in the right column and choose “Banana natural biodiversity mapping”. You will then be asked to answer three more (but mandatory) questions before validating.

1)  Tentative classification: if you have an even slight idea of the classification of the plant, you may add it here. But don’t worry, “other” and “don’t know” are also available!
2)  Environment: a quick indication of where the picture has been shot.
3)  Your appreciation of the health of the plant.
It’s done! Using the mobile application works nearly the same way, but is even easier since GPS coordinates are automatically added from your device (don’t forget to allow localization!).

For more information, please contact:
Christophe Jenny (CIRAD) – christophe.jenny@ or Rachel Chase (Bioversity International) –

MusaTab is an Android© application for recording Musa characterization data in the field. It was developed by Bioversity International for an EU funded project led by the Centre Africain de Recherche sur Bananes et Plantains (CARBAP).

It contains a list of descriptors from the Descriptors for Banana booklet (IPGRI-INIBAP/CIRAD 1996) which can be tailored depending on user preferences. It allows data to  be scored in the field directly on a mobile device and then later uploaded to a computer. It has been tested by the participants of three MusaNet regional workshops in CameroonUganda and Malaysia and is continually being improved based on valuable user feedback.

The minimum requirements to run the application is a mobile device with a screen of 10 inches running an Android™ Operating System from at least version 3. The application needs 23 Mb free space to install. The application is available here (your device will ask you to allow install of the package from an untrusted source). Temporarily allow the install otherwise the app won’t install.

The application works with two types of JSON files, which are not included in the package and must be copied on a specific folder of the tablet. An example of the two required JSON files are availabe in this zipped file. The first file contains the list of accessions to be observed, the second file contains the list of descriptors to score. In addition to these two JSON files, the application requires photos to illustrate some descriptors modalities. These photos are available in this other zipped file (87 Mb).

The user’s manual for MusaTab can be found here.

For any questions, please contact Max Ruas at m.ruas@

MusaID is a taxonomy decision making application that helps users identify an unknown taxonomic unit (taxon) by comparison with a reference collection of known Musa taxa described by a set of qualitative descriptors. It is designed to assist researchers with identification, and it is also a useful working tool for taxonomists themselves. It was first developed by CIRAD in 1990, and recently was improved by a team from CIRAD and Bioversity International.

Taxa are identified through an interactive step-by-step process. A descriptor is selected by the user, or it is proposed by the system in order to optimize the identification sequence by minimizing the number of characters required. The progress of an identification session is evaluated by likelihood measurements, indicating the extent to which an unknown taxon is identical to taxa of the reference collection. 

MusaID relies on morpho-taxonomic data from several collections managed in the Musa Germplasm Infrormation System (MGIS). Descriptors used by MGIS and MusaID are published in “Descriptors for banana (Musa spp.)“, which was jointly produced by CIRAD, INIBAP and IPGRI (now Bioversity International) in 1996. 

The MusaID application is available here: The application does not need to be downloaded and is accessible from any web browser from any platform (Windows, Mac OS, Linux, Android, iOs). Any computer or tablet able to connect to internet can access and use it. The technology used allows it to work in online mode as well as in offline mode (i.e. without any internet connection). To be able to work in offline mode it is necessary to “visit” (to access) all the pages of the web application when online (i.e. connected to internet). 

Members of the MusNet ETG met in December 2017 to discuss and update the current evaluation protocols for three key traits: Fusarium Wilt, Black Leaf Streak and Drought. The publication, Practical Guidelines for Early Screening and Field Evaluation of Banana against Fusarium Wilt, Pseudocercospora Leaf Spots and Drought, were published in 2021.

The team:
Fusarium Wilt – Miguel Dita (EMBRAPA, Brazil), Sijun Zheng (Bioversity International, China), Chunyun Li (Guangdong Academy of Agricultural Sciences, China)

Black Leaf Streak – Luis Perez Vincente (INISAV, Cuba), Francoise Carreel (CIRAD, France), Catherine Abadie (CIRAD, Guadeloupe)

Drought  – Sebastien Carpentier (Bioversity International, Belgium) and Iyyakutty Ravi (NRCB, India

Freedom from viruses is a critical component in the safe exchange of banana germplasm. Material entering the Bioversity International Musa Transit Centre (ITC) is routinely indexed for viruses by collaborating laboratories using a standard protocol including IC-PCR, electron microscopy and symptom observation. These and other assays are described in, the 3rd edition of the “Technical guidelines for the safe movement of Musa germplasm”  published by Bioversity International.

To check the efficiency of testing methods and address aspects that need attention, the MusaNet Conservation Thematic Group (CTG) is facilitating a virus ring test on banana virus detection at participating national laboratories. The test is coordinated by Kathy Crew (Queensland Department of Agriculture and Fisheries) and John Thomas (The University of Queensland). Participants (see list below) were invited to test the routine assays used by Bioversity International, or alternative assays of their choice.

The objectives of the virus ring test are twofold. Firstly, to strengthen the capacity of national institutes to test their material for viruses and secondly, to increase efficiencies by enabling these institutes to carry out their own indexing before sending their material to the ITC for safety duplication and cryopreservation.

The methodology entails blind tests on up to 15 lyophilised leaf samples infected with a defined range of banana viruses, specifically Banana bunchy top virus (BBTV), Cucumber mosaic virus (CMV), Banana bract mosaic virus (BBrMV), Banana mild mosaic virus (BanMMV), and Banana streak viruses (BSV).

For laboratories using the standard protocols, methods and specialised indexing reagents (antibodies and primers) were provided along with the test samples. Positive controls were also provided. Laboratories using alternative assays were requested to provide a summary of their protocol prior to sample dispatch. Individual labs were informed of their own results, but overall test results are presented anonymously. 

The results have been published in a report that can be downloaded here.

For more information on the virus ring test, including if your institute would like to take part, please contact Kathy Crew at Kathy.Crew@