Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Allemagne → 2017 → Acoustic detection of insect pests of stored grains in Kenya

University of Kassel (2017)

Acoustic detection of insect pests of stored grains in Kenya

Njoroge, Anastasia Wanjiru

Titre : Acoustic detection of insect pests of stored grains in Kenya

Auteur : Njoroge, Anastasia Wanjiru

Université de soutenance : University of Kassel

Grade : Doktor der Agrarwissenschaften (Dr. agr.) 2017

Kenya experiences persistent food and nutritional security just like other countries in SubSaharan Africa (SSA). This is aggravated by low productivity, erratic weather patterns and high postharvest losses ranging from 20-30%. Insect pests are a key constraint to effective preservation of grain crops in sub-Saharan Africa (SSA), with severe damage caused by these pests during bulk storage in warehouses. Although a number of approaches have been advanced for control of storage pests, their elimination is still yet to be realized. Maize and common beans are Kenya‘s main staple food crops with cereal crops playing a major role in smallholder farmers’ livelihoods. To achieve food security and food safety several pest management approaches including the use of pesticides are employed have been advanced for management of post-harvest pests. The major postharvest pests in Kenya are Prostephanus truncatus Horn (Coleoptera : Bostrichidae), Sitophilus zeamais Motschulsy (Coleoptera : Curculionidae), Tribolium castaneum Herbst (Coleoptera : Tenebrionidae) for maize and Acanthoscelides obtectus Say (Coleoptera : Bruchidae) for common beans. Their presence often goes unnoticed in freshly harvested grain or during grain procurement in storage warehouses. Their early detection is a challenge because conventional sampling methods cannot detect infestations of less than 5 insects/kg. Based on this, a study was conceived to study the acoustic characteristics of these postharvest pests in the laboratory as well as in bulk storage warehouses. State-of-the-art acoustic measurement equipment and non-pheromone grain probe traps were used for the study. Laboratory studies focussed on the possibility to distinguish between different species as well as between larval and adult acoustic emissions. In the storage warehouses an acoustic survey was done to distinguish the detection range of two acoustic devices as well as compare acoustic data with counts of insects captured using traps. Lab results showed that the three insect species could be distinguished based on their spectral profiles which ranged from Broadband frequency with multiple peaks from 1-8 kHz, HighFrequency with a band of high energy between 4 and 5.5 kHz, Mid-Frequency1 with a peak near 3 kHz and a smaller peak between 190 3.5 and 5.2 kHz, Mid-Frequency 2 with a band at 3 kHz, and Low-Frequency with a peak between 500-700 Hz. Majority of the insect sounds were between 3 - 8kHz. The differences in their acoustic signals can be harnessed in designing sensors within their frequency range. Their frequency range of detection was identified to be between 3.5 and 5.5 kHz. Field results from the acoustic survey showed that in their natural habitat postharvest insects existed as mixed species and there were at least two species present in each store that was sampled. There was an economic incentive to fumigate all the stores we visited in response to the presence of these pests mainly because of they are internal infesting species. Background noise was a major challenge in the detection of the infestation especially because most stores were located in urban and peri-urban areas with high human activity. The convergence of the acoustic signal analysis and the captured insects demonstrated that indeed acoustic detection is a useful tool in explaining infestations in grain stores in Africa. In conclusion, protection of strategic grain reserves can help reduce future food shortages. Since acoustic technology can be incorporated in pest surveillance programs in our major warehouses, further research into affordable acoustic sensor development is recommended


Version intégrale (2,22 Mb)

Page publiée le 18 janvier 2019