*2.2. Storage Technologies*

Four types of storage structures were used to store maize grains during the experiments: (1) Polypropylene bags with a capacity of 100 kg; (2) Purdue Improved Crop Storage (PICS) bags with a capacity of 100 kg; (3) Improved clay granaries of different capacities, the largest of which was 1000 kg; (4) Metal silos with a capacity varying from 250 to 1000 kg. Generally, in both study areas, farmers had been using improved clay granaries and polypropylene bags for grain storage Sissinto-Gbénou [10]. However, as the polypropylene bag was a widely used storage structure, it was considered as the control structure in the experiment. The other storage structures offered alternatives to producers. The polypropylene bag and the improved clay granary were storage technologies that already exist in rural areas. The PICS bag was developed by Purdue University and has been more commonly used to store cowpea than maize. Both the large metal silo and PICS bag were introduced in Benin for product storage at a community level. In this study, we used silos for agricultural product storage at an individual producer level.

In addition, maize was treated using Actellic® Super powder (2% pirimiphos-methyl) to protect the stocks of maize against pest attacks. The insecticide was applied at a rate of 50 g Actellic® Super powder per 100 kg maize grains. This insecticide is recommended by the Plant Protection and Phytosanitary Control Service to farmers in the Republic of Benin. It is an easily biodegradable insecticide that is only slightly persistent. Actellic® Super is effective in combating *Sitophilus zeamaïs*, but is ineffective against borers (Bostrichidae), including the grain borer (*Prostephanus truncatus*). It is sold under the trade name of Actellic® Super.

#### *2.3. Vegetal Material*

The grains of the improved maize variety DMR were used for the experiments in both areas. They were provided by the farmers involved in the experimental trials. The grains had been sorted and dried before storage. The grain moisture levels were recorded at the beginning of the trials.

### *2.4. Experimental Setup and Treatments Implemented*

Height treatments were designed for the experiments. Each treatment was a combination of one storage structure, with and without the application of Actellic® Super powder. The experiment was arranged in a completely randomized block design, with a 2\*4 factorial experiment for the first year of experiments and a 2 \* 5 factorial experiment

for the last two years, including the metal silo in the trials. In addition, the principle of divided plots (split-plot) was used. The experience plan was unbalanced because the treatments did not have the same number of repetitions. Two categories of experimental farmers were defined depending on the application of Actellic® Super powder to protect the stocks of maize against pest attacks. The treatments were as follows:


#### *2.5. Sampling and Data Collection*

During the three experiments years, technical and economic data were collected each year at monthly intervals for seven months, from the establishment of the experiments to the maize destocking. Data collection covered the period from February or March to August or September each year. This period coincided with the seven or eight months of maize storage each year. Each month, four samples of 750 g per treatment (one per treatment) were taken from each experimental producer for damage and loss evaluation of the storage treatments in the laboratory. A one-meter handle ladle was used to take samples by mixing the grains in the stock enclosure before the sample was taken. Each sample was then saved in an envelope, which was packed in a labeled plastic bag and sealed. The samples were then transported to the laboratory of the Agricultural and Food Technology Program (PTAA) of the National Institute of Agricultural Research of Benin (INRAB) in Porto-Novo for further analysis. In the laboratory, the samples were stored in a refrigerator at a temperature of 4 ◦C for the two weeks required for the loss rate assessment.

Data on maize storage costs (from harvest to storage) were collected from each trial participant. They included the cost of construction or purchase price of the storage structures, the cost of labor related to maize storage operations (including dispatching and shelling costs), and the cost of the chemical for storing the maize in the structures. Selling prices per kilogram of maize were collected per month for the duration of the experiments at the level of each participant. The interest rates charged by the microfinance institutions (10% and 12%, in Savalou and Boukoumbé, respectively) and the lifespan of the storage structures (1 year for the polypropylene bag, 2 years for the PICS bag, 20 years for the metal silo and 15 years for the improved granary) were considered for the calculation of storage costs. A structured questionnaire was designed to collect these data. The results of the evaluation of the loss rates were used to calculate the amount of financial loss for each storage treatment.
