SATrends ISSUE 38                                                                                                                  January 2004

  • Pest Population Dynamics
  • More crops per drop
  • SAT data from satellites
  • Des chercheurs …de poussières
  • 1. Pest Population Dynamics
    top 

    Insect pests are a major cause of crop loss globally.   Pest management will be effective and efficient if we can predict the occurrence of peak activities of a given pest. Research efforts are going on to understand the pest dynamics by applying analytical and other techniques on pest surveillance data sets.

    One of the problems in addressing pest management is inadequate knowledge about the factors influencing pest population dynamics. To understand pest dynamics, scientists collect pest surveillance data and details of pest incidence, climatic, soil, and agricultural practices. Correlations between some of these factors and pest incidence, based on statistical models were developed to predict when to apply pest management instruments. However, a functionally viable model that can be used by the farmers is not yet available.

    The Helicoverpa pest feeding on a chickpea plant
    The pod borer, Helicoverpa armigera is one of the key pests causing severe yield losses, infesting several crops such as cereals, pulses, cotton, vegetables and fruit crops as well as wild and weedy hosts of crop plants. The non-linear and complex nature of Helicoverpa population dynamics makes it difficult to predict population densities using traditional forecasting models. An effort has been made to understand the Helicoverpa population dynamics on the chickpea crop using modern information and communication technologies (ICT), such as neural networks, to analyze and interpret long-term data. ICRISAT has collected daily and weekly recordings about weather and pest incidence at various locations in the farm for 25 years.

    A neural network is an interconnected set of input/output units where each connection has a weight associated with it. They can be “trained” to recognize a pattern. During the learning phase, the network learns by adjusting the weights so as to predict the call label of input samples during the testing phase. Neural networks possess high tolerance to noisy data as well as the ability to identify patterns on which they were not trained.

    Graduate students from the International Institute of Information Technology (IIIT), Hyderabad, and ICRISAT conducted experiments. The results show that it is possible to predict a pest attack with high probability for one week in advance. These pest predictions could help the farmers in pest management programs with improved environment quality, as it can avoid misuse of chemical pesticides. This forecast model is currently under validation during the present cropping season at ICRISAT-Patancheru.

    For more information, contact g.rangarao@cgiar.org or c.gowda@cgiar.org

    2. More crops per drop
    top 

    ICRISAT and USAID have worked together for many years, fighting poverty in drought-prone areas in Africa. Their latest partnership, under the LEAD project (Linkages for the Economic Advancement of the Disadvantaged) is a good example of how much can be achieved using simple, readily available technology.

    Drip irrigation is a highly efficient way to use water. Expensive, high-tech equipment is available, but so are simple, basic “drip kits”, ideal for smallholder farmers. The kit is essentially a plastic water container and a series of pipes ending in specially designed miniature “taps” through which the water is delivered, drop by drop. Result? The water gets directly to the plants, with no wastage. The system can be used for any crop. But obviously it’s most profitable to use it on cash crops, such as vegetables.

    The program targets three drought-prone districts in Zimbabwe: Chivi, Zaka, and Tsholotsho. The first kits arrived in July 2003. Today, nearly 1200 kits have been distributed, and 1200 gardens have been set up. Farmers grow a range of crops – tomatoes, onions, okra, varieties of spinach, cabbages, and maize.

    The drip irrigation kit in use

    Even before distributing the kits, the project provides training on how to use them. To date over 2000 farmers have been trained, including 180 “contact farmers” who act as unofficial extension agents for the new technology. The training covered various aspects – installation and maintenance of the kit, establishing and managing a garden, including recommended management practices for different vegetables, control of pests and diseases, and record keeping. Farmers now know how to manage irrigation schedules – eg, how to inspect soil to judge when the next watering is required and how much water to apply.

    About one-third of the farmers have already harvested one crop, and are busy establishing the second crop. Production levels have been astonishing, considering that the project began just 6 months ago. Nearly 85 tons of produce have been harvested. Most of this was used as food for the family – a huge boost for nutrition in traditionally poor, nutritionally deficient communities. Some was also sold for cash – total sales in the past 6 months were Z$ 4.8 million.

    When the project was being designed, we were concerned that adoption of the drip kits would be low because of heavy competition for scarce water supplies. In fact, drip irrigation has become highly popular – simply because for most people, this was the only way they could ever plant a vegetable garden in a drought-prone area.

    For more information, contact e.monyo@cgiar.org

    3. SAT data from satellites
    top 

    Climate research is increasingly becoming important in Semi-Arid Tropics (SAT) agriculture. A sudden dust storm in the desert margins, or long breaks in rainfall during the monsoon result in significantly reduced productivity in SAT farms. There are relatively limited methods available to predict such events. The possibility of using data derived from the weather satellites directly is likely to change this scenario.

    The US National Oceanographic and Atmospheric Administration (NOAA) has several satellites in space for monitoring weather events. Some of these are geo-stationary satellites (they appear to be stationary at one point over the earth because of matching speeds) which provide climate data about a particular region over the earth’s surface. Another class, the polar orbiting satellites, are positioned at about 800 km from the earth and fly over the same location twice in a span of 24 hours. These satellites send data that is rich and complex, and which are potentially useful in making superior weather forecasts when combined with data on suitable ground parameters. Currently, NOAA12 to NOAA15 are the NOAA satellites in polar orbit.

    Recent NOAA satellite image
    Until recently data reception was a difficult and expensive process and was rarely attempted in research institutes. Availability of reasonable computing power at the desktop has changed this scene. It is now possible to build or buy a simple antenna with a PC-compatible card that can work with a Windows-based PC. The NOAA signals are powerful enough to be captured with a simple crossed dipole antenna, and can be converted into visual signals using off-the-shelf software. They can also be captured as “raw” data that can be manipulated easily using standard techniques in database management. The whole setup costs about $ 600 (excluding the PC costs). One such arrangement has been commissioned in ICRISAT-Patancheru by the ISU and has been functional for the last three months. The setup took less than two hours and was easy. Over the period of three months the ISU team has upgraded the software, and the captured images (six maximum per day from three NOAA satellites) which are accessible from the Institute Intranet. Efforts are on to use such data in developing a local climate prediction model. ICRISAT scientists are in touch with the NOAA to explore further possibilities and the use of superior sensors on the ground.

    For more information, contact s.walia@cgiar.org or r.kumar@cgiar.org

    4. Des chercheurs …de poussières
    top 

    Thomas Maurer non seulement savoure le Niger mais en tire le maximum. Il profite de sa présence ici pour écumer les routes, prospecter et dénicher les meilleurs sites, y faire tourner sa machine à vent, un énorme tunnel aspirant avec lequel il secoue la terre pour récolter particules et poussières, qu’il démontera sitôt l’expérience finie pour repartir encore sur les routes et recommencer toute l’opération. Il a trouvé une aide précieuse dans son assistant Tahirou Boye qui est lui aussi totalement hermétique à la fatigue. La machine à vent n’est pas légère (voir schéma en fin d’article), ni facilement transportable, les routes et pistes nigériennes n’offrent pas non plus les meilleures conditions de travail. Et à les voir tous deux revenant d’excursion, toujours heureux, déchargeant leur volumineux engin et par la suite de toutes petites sacoches remplis de poudre, tout un chacun se demanderait s’il n’aurait pas affaire à deux chercheurs d’or.

    SATrends a donc légitimement voulu savoir où ces deux chercheurs prospectent au cas où le filon pourrait être partager. Mais voilà la réponse obtenue:

    L’explication de cette surprenante réponse est que Thomas Maurer ne se trouve pas totalement au Niger mais bien plus à l’intérieur d’un grand programme allemand de recherche climatique, le DEKLIM. Il est doctorant de l’institut de pédologie de l’Université de Hohenheim (UH) dont l’ICRISAT-Niamey est un institut collaborateur. Thomas a pour mission de mesurer les émissions de poussières en condition semi-réelle.

    Thomas avec la machine à vent

    Donc cette poudre qui remplit ces sacoches, l’or de Thomas Maurer, n’est rien d’autre que de la poussière du Sahel, celle qui, véhiculée par les vents et diffusée dans l’atmosphère planétaire, rentre dans les énormes équations climatiques que le DEKLIM cherche à maîtriser. C’est en multipliant les sites et les prélèvements que Thomas Maurer affinera son étude. Une fois son bureau de Hohenheim réintégré, il extrapolera ses données à l’aide d’images satellite du Sahel. Il saura y reconnaître les différents types de sol et caractériser leur production de poussière en fonction de la gamme des vents.

    En guise de conclusion à notre interview, Thomas nous offre le schéma de ce simulateur, dont il est même le concepteur. Sa machine a été homologuée par les meilleurs spécialistes en la matière, notamment ses aînés américains issus d’une longue lignée de chercheurs engendrée par les dust bowl des années 1930.

    Alors peut-être qu’avec ce schéma, nos lecteurs pourront eux aussi se transformer en chercheurs…de poussière.

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

     

             

          

     

    Pour plus d’information, contacter m.maurer@cgiar.org

    top