AFROweeds and Weedsbook: Online and offline weed identification and management support for better-informed agricultural change agents

Weed identification is critical for generating suitable and targeted weed management recommendations for farmers. However, accurate and timely identification requires specialized botanical knowledge and skills, which are rarely available among the agricultural change agents working in the field. With the primary aim of assisting those agents in communicating and helping rice farmers with specific weed problems in their field, a team of weed scientists and botanists from CIRAD and AfricaRice, in collaboration with national agricultural research system scientists and agricultural change agents from 13 African countries, developed an interactive electronic tool called 'AFROweeds' and an online exchange platform called 'Weedsbook'. The IDAO AFROweeds identification tool is a computer program (also called identikit) supported by a database that can be accessed online (www.afroweeds.org/idao), or used offline as a program on a desktop or laptop computer, or as an app on a smartphone or e-tablet. The identikit enables the user to select some of the most eye-catching characteristics of the plant to be identified and to indicate what the character looks like using pictorial multiple-choice menus. The identikit computes the probability of the species identity corresponding to the combination of choices. The user can view a list of species with percentages of fit with the selected character shape, forms or colors and check the most likely species (e.g. those with 100% fit) with the actual plant. Clicking on the species names in the list provides access to species' pages from the database, with information on ecology, biology and management. Users can also compare the plant being examined with field and herbarium photos or illustrations. The tool currently contains close to 200 species, primarily for lowland rice, but will be upgraded with species from other rice-growing environments. The AFROweeds tool was tested on an electronic tablet in the field in Zoungo in the Oueme Valley in Benin (07°06'46" N; 02°30'58" E) with 24 weed scientists, botanists and agricultural change agents. Each test was carried out by two individuals at a time. The average identification time was 6 minutes 34 seconds, ranging from 1 min 14 s to 10 min 16 s, with 12 successful identifications out of 16 attempts (75%). Identifications were unsuccessful when the species was not included in the database, or when the user made an unnoticed early mistake in the selection procedure. Alongside the development of the AFROweeds identification tool, the online African weed science network Weedsbook has been established. This open-access platform enables professionals to exchange information or request assistance on all aspects of weeds and weed management. Farmers can benefit from the tool and network through interaction with better-informed and better-equipped agricultural change agents. Improved support to farmer decision-making should lead to better weed management and lower weed-inflicted yield losses, and subsequently contribute to improved food security. (Resume d'auteur)