^{1}

^{*}

^{1}

^{*}

^{1}

^{*}

^{2}

^{*}

In Senegal, millet ( Pennisetum glaucum (L.) R. Br.) and rice ( Oryza sativa (L., 1753)) are the most widely consumed foods. This study is part of improving the conservation of these two cereals in Senegal by assessing the quantitative and qualitative losses caused by a lepidopteran, Corcyra cephalonica (Stainton) subservient to millet and rice stocks. For this purpose, samples of millet and rice from an area of the center of the groundnut basin (Diourbel) were collected, sterilized in the cold and infected with C. cephalonica eggs from the same locality. These infected samples were tracked during a development period of two successive generations. The samples were scrutinized before being infected and after a larval cycle of codling moth. The results showed that rice grains are richer in water (10.75% ± 0.4249%, on average) than millet (9.40% ± 0.3944%, on average) and the difference in rank is very significant (p-value = 0.0001 < 0.05). Moreover, the attack percentage on millet grains is three times higher (36.31% ± 25.18%) than rice (12.95% ± 6.69%) with a non-significant difference (p-value = 0.296 > 0.05 ). A similar trend is observed at the loss percentage, which is four times higher with millet grains (8.67 % ± 5.07 % ) than rice (2.86 % ± 2.75 % ) with a non-significant difference (p-value = 0.835 > 0.05). A multiple linear regression showed a generation effect on millet for the attack percentage and a generation and cereal effect for the percentage of weight loss on rice.

Losses occur at all stages, from harvesting to consumption: first at the producer, whether they are products for self-consumption, seed or waiting to be marketed; then during transport to storage locations and during storage; finally, in the reserves of traders. All these losses are all the more damaging since they affect the product once harvested; they do not only harm the peasant but also cost very much to the national economy [

The study was done in the Laboratory of Life and Earth Sciences of the New High School of Kaolack (NLK). This environment belongs to the Sahel-Sudan area (SAS) and the experiments were made under an average temperature of 30.44 ± 1.96 degrees Celsius and an average relative humidity of 73.21% ± 1.74%. Samples of millet and rice were collected from the seed packing station in the Diourbel region.

The biological material consisted of samples of millet and white rice, eggs of C. cephalonica recuperated from the clutches of females from these two cereals from the same locality. The physical material was formed by the following: plastic boxes with holed lids to facilitate the aeration of the mid 50 cm^{3} capacity, mating drums for egg recapture, bins to hold samples, a storage cabinet to protect samples from rodents, a refrigerator for the sterilization of the food substrate, a column of sieve to separate the moths from the cereal, mixed dishes and a clamp for sorting grains and assisted by a binocular magnifying glass, an electronic scale Jeulin: cap: 400 g; grad: 0.1 g for weighing samples, 80 ml Pyrex beakers that resist heat during heat, a test for determining the water content of samples, labels for marking samples, a GPS (Global Positioning System) for the survey of geographical coordinates.

Samples of 20 g millet grains or white broken rice, infested with eggs from a wild strain of C. cephalonica from the seed packing station in Diourbel, located at 14˚39'4.5''N; 16˚15'19.36''W belonging to the Sahel-Sudan zone (SAS), used to study the biological parameters of dry grain moth, are collected after a first and second full cycle of larval moth development and have been used as biological material. Samples were carefully examined before contamination and after the larval moth cycle.

A decrease in the physical substance of the product results in weight loss. However, weight loss and product loss should be distinguished. The decrease in moisture content leads to a decrease in weight, which is not a food loss. In opposition, an increase in weight by absorbing moisture, as a result of rains on an open-air stock, for example, can cause serious damage that will result in losses [

Thus, to determine the water content of cereal grains, two batches of ten boxes each are recuperated. The contents of the boxes in Lot 1 are mixed to form an overall sample of 200 g of millet grains on the one hand, those of Lot 2 are also mixed to give an overall sample of 200 g of rice grains on the other.

These global samples are then sifted through a column of decreasing mesh sieve to eliminate insects in their various forms and dust. Global samples cleaned, ten Pyrex glass beakers with 80 ml capacity are each filled with 20 g of millet grains and ten more of 20 g of rice each. These beakers are then placed in the oven at a temperature of 85˚C. After 17 hours, the samples are weighed again. The water content was assessed by the following formula:

Te ( % ) = m 1 − m 2 m 1 ∗ 100

with: Te: water content; m1: sample mass before deposition in the oven; m2: sample mass at the end of the oven.

The term “post-harvest loss” means a measurable quantitative and qualitative reduction of a given product. These losses can occur throughout the various phases of the post-harvest system. This definition must also take into account cases of product deterioration. However, it would be more accurate to speak of limitation in the use of the product than losses themselves. Grains partially damaged by insects, for example, may no longer be suitable for human consumption, or for marketing. If such were their destinations, we must admit that these are losses, even if grain can be recuperated by using them for the feeding of farm animals [

The percentage of grain attacks (% A);

The percentage of grain weight loss (% B).

Based on the data collected, the percentages of grain attacks were calculated after sieving using the following formula:

% A = Nb .GA Nb .GA + Nb .GS ∗ 100

with: Nb.GA: the number of grains attacked; Nb.GS: the number of healthy grains.

Weight loss (expressed as %) were calculated after each sifting from [

% B = PGS ∗ Nb .GA − PGA ∗ NGS PGS ( Nb .GA + Nb .GS ) ∗ 100

with: Nb.GA: the number of grains attacked; Nb.GS: the number of healthy grains; PGA: the weight of the grains attacked; PGS: the weight of healthy grains.

The type of demonstrating used here is a multiple linear regression with Y/variables depending on the attack percentage (% A) and the percentage of loss (% B) and X/variables explaining the number of well grains (Nb.GS), the number of attacked grains (NB.GA), the weight of well grains (PGS), the weight of the attacked grains (PGA), the type of grain (mil and rice) and the larval generations (G0, G1, and (G1 + G2)). Level two coactions between the factors were used in this regression with Excel and XLSTAT, which enabled the Calculations and comparison tests of the Mann Withney type as well as the ANCOVA. Square root transformation [

The water content of millet and rice grain samples was determined and the results showed that the highest water content was obtained with rice grains ranging from a minimum of 10% to a maximum of 11.50% and an average of 10.75% ± 0.4249%, and the lowest on millet grains ranged from 8.5% to 10% with an average of 9.40% ± 0.3944%. Thus the grains of rice are richer in water than those of millet and the difference in rank is very significant (p-value = 0.0001 < 0.05).

For all two cereals (mil and rice) and generations, the attack percentage ranges from a low of 3.83 to a high of 58.74% with an average of 24.6263% ± 4.8208%. The percentage of loss ranges from a minimum of 0.18 to a maximum of 13.37% with an average of 5.7275% ± 4.8202%. Thus the percentage of loss is less than the percentage of attack and its evolution follows the percentage of attack.

The number of good grains ranged from a minimum of 581 to a maximum of 1503 grains with an average of 913 ± 375.6914 grains; the number of grains attacked fluctuates between 24 and 1028 grains with an average of 377.375 ± 409.4808 grains, this gap indicates that the data are widely dispersed around the average. This seems to be normal since cereals do not have the same nature and then grains are attacked randomly.

The weight of good grains ranges from 4.5 to 9.6 g with an average of 7.9 ± 2.0142 g; the attacked grain is between 0.4 and 5 g with an average of 2.0375 ± 1.84 g (

For the type of cereal, we have the same modality (50% everywhere). However, the modality is in favour of the generations (G1 − G2) (

The percentage attack is three times higher on millet grains than on rice, with 36.31% ± 25.18% and 12.95% ± 6.69% respectively. However, the difference in degree between the percentages of attacks on millet and rice is not significant (p-value = 0.1936 > 0.05).

Results for the cereal-mil group:

The results showed that there is no relationship between the percentages of attack and loss in millet grains (Sig. = 0.2 > 0.05).

Results for the cereal-rice group:

The results showed that there is no relationship between the percentages of attack and loss in the grains of rice (Sig. = 0.2 > 0.05).

Variable | Observations | Minimum | Maximum | Average | Standard deviation |
---|---|---|---|---|---|

% A | 8 | 3,83 | 58.4 | 24.6263 | 21.1382 |

% B | 8 | 0,18 | 13.37 | 5.7275 | 4.8208 |

Nb.GS | 8 | 581 | 1503 | 913 | 375.6914 |

Nb.GA | 8 | 24 | 1028 | 377.375 | 409.4808 |

PGS (g) | 8 | 4.5 | 9.6 | 7.9 | 2.0142 |

PGA (g) | 8 | 0,4 | 5 | 2.0375 | 1.84 |

Variable | Terms | Staff | % |
---|---|---|---|

Cereal | Millet | 4 | 50 |

Rice | 4 | 50 | |

Generation | G0 | 2 | 25 |

G1 | 2 | 25 | |

G1 + G2 | 4 | 50 |

The best model for the chosen selection criterion is shown in bold (

The effects are detailed in

The standardized coefficient graph (

Variables | Variables | MCE | R^{2} | Adjusted R^{2} | AIC of Akaike | SBC of Schwarz |
---|---|---|---|---|---|---|

2 | Nb.GA * G/Cereal * G | 0.0006 | 1 | 1 | −62.6764 | −62.1203 |

The best model for the chosen selection criterion is shown in blue. Caption: G = Generation; Nb.GA = Number of Attacked Grains, Interaction.

Source | DDL | Sum of squares | Average squares | F | Pr > F |
---|---|---|---|---|---|

Nb.GA * Generation | 3 | 136.4464 | 45.4821 | 82654.7274 | 0.0026 |

Cereal * Generation | 3 | 41.7398 | 13.9133 | 25284.6099 | 0.0046 |

Caption: Nb.GA = Number of Grains Attacked, * = “Co-action”.

Value | Standard error | t | Pr > |t| | Lower bound (95%) | Upper terminal (95%) | |
---|---|---|---|---|---|---|

Constant | −0.0718 | 0.0779 | −0.9225 | 0.5257 | −1.0610 | 0.9174 |

Nb.GA * G0 | 0.1626 | 0.0034 | 47.9863 | 0.0133 | 0.1195 | 0.2056 |

Nb.GA * G1 | 0.0806 | 0.0004 | 191.2634 | 0.0033 | 0.0752 | 0.0859 |

Nb.GA * (G1 + G2) | 0.1388 | 0.0006 | 214.3765 | 0.0030 | 0.1306 | 0.1470 |

Cereal-mil * G0 | −15.3471 | 0.4382 | −35.0196 | 0.0182 | −20.9155 | −9.7787 |

Cereal-mil * G1 | −9.4754 | 0.0895 | −105.8220 | 0.0060 | −10.6132 | −8.3377 |

Cereal-mil * (G1 + G2) | −83.8396 | 0.5850 | −143.3039 | 0.0044 | −91.2734 | −76.4059 |

Caption: G0 = Generation 0; G1 = Generation 1; G2 = Generation 2; Nb.GA = Number of Grains Attacked, * = “interaction”.

In short, we can say that there is an effect of larval generations on the number of attacked grains and that is positive and significantly differ on the percentage of attack; an effect of larval generations on millet cereal that is negative and significantly different on the percentage of attack.

Analysis of

Graphiques des moyennes:

The percentage of loss is four times higher in millet grains than in rice, 8.67% ± 5.07% and 2.86% ± 2.75% respectively. They are not correlated (Sig. = 0.8 > 0.05).

The best model for the chosen selection criterion is shown in bold (

Variables | Variables | MCE | R^{2} | Adjusted R^{2} | AIC of Akaike | SBC of Schwarz |
---|---|---|---|---|---|---|

2 | PGS (g) * PGA (g)/Cereal * Gen | 0.0002 | 1 | 1 | −71.1163 | −70.5602 |

The best model for the chosen selection criterion is shown in bold. Caption: PGS = Healthy Grain Weight; Gen = Generation; PGA = Weight of Attacked Grains, * = co-action.

Source | DDL | Sum of squares | Average squares | F | Pr > F |
---|---|---|---|---|---|

PGS (g) * PGA (g) | 1 | 138.7215 | 138.7215 | 724,014.4724 | 0.0007 |

Cereal * Generation | 5 | 23.9617 | 4.7923 | 25,012.1080 | 0.0048 |

Caption: PGS—Healthy Grain Weight; PGA—Weight of Attacked Grains, * = co-action.

The model parameter table (

The standardized coefficient graph (

Average charts:

The analysis in

Source | Value | Standard error | t | Pr > |t| | Lower bound (95%) | Upper terminal (95%) |
---|---|---|---|---|---|---|

Constant | −11.1734 | 0.0756 | −147.7179 | 0.0043 | −12.1345 | −10.2123 |

PGS (g) * PGA (g) | 1.1542 | 0.0063 | 183.4351 | 0.0035 | 1.0743 | 1.2342 |

cereal-mil * G0 | 5.9094 | 0.0380 | 155.3909 | 0.0041 | 5.4262 | 6.3926 |

cereal-mil * G1 | 3.0876 | 0.0188 | 164.5059 | 0.0039 | 2.8491 | 3.3261 |

cereal-mil * (G1 + G2) | −3.1782 | 0.0727 | −43.7441 | 0.0146 | −4.1013 | −2.2550 |

cereal-rice * G0 | 6.9212 | 0.0536 | 129.1414 | 0.0049 | 6.2402 | 7.6021 |

cereal-rice * G1 | 6.8753 | 0.0250 | 275.0169 | 0.0023 | 6.5577 | 7.1930 |

Legend: PSG = Weight of Healthy Grains; G0 = Generation 0; G1 = Generation 1; G2 = Generation 2; PGA = Weight of Attacked Grains, * = co-action.

On rice cereals, on the other hand, this average is higher in the G1 generation (6.09%), followed by that of the generation (G1 + G2) (2.59%). and lower with generation G0 (0.18%) (see

Thus, we can talk about a single generation effect on millet cereal and a cereal and generation effect on rice cereal.

Rice (Oryza sativa L.) and millet (Pennisetum typhoid Stapf. and Hubb.), stored are attacked by a wide range of insects whose identities and modes of attack differ from those encountered in the field. These are mostly polyphage insects attacking not only these two grains in stock, but also other stored grains. Among these insect pests, Corcyra cephalonica is arguably one of the most formidable that can result in both significant quantitative and qualitative losses during storage. [

Of our results, the water content (Te) is higher on rice grains (10.75% ± 0.4249%) millet (9.40% ± 0.3944%) with a very significant difference (p-value = 0.0001 < 0.05). These values are high and can result in considerable losses during storage.

According to [

Our results showed that the attack percentage can vary from 3.83% to 58.74% and the percentage of loss from 0.18% to 13.37% for both cereals.

The percentage of attack is three times higher on millet grains (36.31% ± 25.18%) than in rice (12.95% ± 6.69%) with a non-significant difference (p-value = 0.296 > 0.05). A similar trend is observed in the percentage of loss, which is four times higher in millet grains (8.67% ± 5.07%) than rice (2.86% ± 2.75%) with a non-significant difference (p-value = 0.835 > 0.05). This would be due to the pounding operations carried out on millet grain and which would weaken this cereal so that its grain still retains its envelopes. As the white rice grain is polished and smooth, this texture would probably cause the lower percentages of attack and loss for this cereal. The results of [

In additional, the percentages of attack and loss in the grains are not correlated.

The results of the ANCOVA show that the coactions of the factors number of grains attacked and the larval generations on the one hand, the type of cereal and the larval generations on the other, have significantly different influences on the percentage of attack of C. cephalonica on millet and rice grains.

Indeed, interactions of the number of grains attacked with larval generations have positive effects, while those of millet cereal with these same generations have negative effects on the percentage of attack of C. cephalonica.

The effect of cereal-generation interdependence on the attack percentage is constant and zero on rice cereal. However, its evolution is not linear and shows a mass effect that occurred during the second generation in millet grains. This allows us to say that the effect of cereal-generation interdependence on the attack percentage evolves differently in millet cereal. We can talk about a generation effect on millet for the attack percentage.

From our results, the coactions of factors: the weight of healthy grains with the weight of the attacked grains on the one hand, the generations with the type of cereal on the other hand, have significantly different influences on the percentage of weight loss in dry matter caused by C. cephalonica. These influences are positive with the exception of the factor cereal-mil * (G1 + G2) which has a negative influence.

This negative aspect of this source on %B would likely be due to overcrowding in the second generation, as we have observed for the attack percentage. This same effect is recorded during the passage between the G1 and (G1 + G2) generations in the rice grains.

The effects of cereal * Generation interdependence on the percentage of loss in millet cereals are increasingly and linearly changing, and differently in rice grains. These observations allow us to talk, in terms of the percentage of loss, of a single generation effect on millet cereal and a grain and generation effect in rice cereal.

Two types of effects were noted in this study: a generation effect on millet for the attack percentage and a generation and cereal effect for the percentage of weight loss on rice. Particular attention should be paid to the affinity of the moth of the grains with millet grains, where the attack percentage and weight loss are higher. The study also showed that long-term conservations increase the generations of the pest as well as the effect of generational and cereal interaction on the quantitative and qualitative losses of stored foodstuffs, as well as the occurrence of a mass effect on insect populations. A focus on the molecular biology of C. cephalonica would provide a better understanding of the ecology of C. cephalonica with respect to millet and rice food substrates.

The authors declare no conflicts of interest regarding the publication of this paper.

LO, M., Diome, T., Thiaw, C. and Sembene, M. (2021) Evaluation of Quantitative and Qualitative Losses on Millet and White Rice in Storage Grains Caused by Corcyra cephalonica (Stainton) in Senegal. Advances in Entomology, 9, 30-43. https://doi.org/10.4236/ae.2021.91003