Variability of thermal and water conditions inside storage structures.
The highlighting of the thermal and hydrous variants of the seeds during the conservation period, allowed us to realize the graph below. It emerges that there is a thermal variability in the stores which would act significantly on the moisture content of seeds in storage. The thermal gradient and that of moisture of the seeds evolve in opposite directions. So the hotter the store is, the lower is the moisture content of the seeds. This thermal variation would be due to the alternation of diversified periods of sunshine of the environment during the period of storage. This is all the more true since this conservation period generally coincides with the dry season during which high temperaturees are often recorded. The second sampling, which coincides with the sowing period, is characterized by a higher temperature (31.59 ° C) with a relative humidity of 20.16%. This analysis shows that the temperatures and moisture levels of cottonseed stored vary from one store to another and between districts (Fig. 1).
Figure 1: Evolution of temperature and aridity index (Source: field data, 2019)
Variation in germination rate of seeds at the beginning of conservation.
Table 1 below summarizes the probabilities derived from analysis of variance (ANOVA) on seed germination rates from different storage locations. Indeed, it appears that there is a significant difference at the 5% threshold between the seed germination rate (Pvalue = 0.02). Moreover, the Student-Newman-Keuls comparison test reveals a homogeneity of seed germination rate at the threshold of 5%. The average seed germination rate of Fo_Boure, the Municipal Agricultural Development Sectors (SCDA) of N'dali, Sonnoumon and Soka stores are virtually identical. The slight difference is therefore observed only at the level of seeds taken from SCDA Sinendé, Sekere and Yarra stores with a low coefficient of variation (2.42%). Thus, the highest rates are those of Sonnoumon seeds (73.25%) while the lowest are recorded in SEKERE (69.75%). This translates that at the beginning of conservation, the seeds always keep their initial germination rates and know no disturbance of the environment.
Table 1
Variance analysis of seed germination rate at the beginning of storage (Sample1)
Shops | Germination rate(%) |
Fo_Boure | 73 a |
SCDA_N’dali | 73.25 a |
SCDA_Sinendé | 71 a |
Sekere | 69.75 a |
Sokka | 72.5 a |
Sonnoumon | 73.25 a |
Yarra | 70 a |
Moyenne | 71.82 |
Pr | 0.023* |
F | 3.146 |
CV (%) | 2.42 |
F: F of Fisher Pr: probability *: significant at 5%; CV (%): Coefficient of variation. On each line, the averages followed by the same alphabetical letter are not different at the 5% threshold according to the Student-Newman-Keuls test.
Temporal evolution of seed germination rate
Analysis of the variance showed a significant difference at the 5% threshold for the tests (Pr (F) = 0.001), this confirms that the average seed germination rate varies according to the stores. So, seed storage conditions usually vary from one store to another and over time. The seeds are preserved for this purpose in a diversified way. This reflects a likely and specific influence at each store according to the conditions imposed on the seeds. The associated comparison test stipulates a strong heterogeneity within the stores. The lowest germination rates are recorded in the third sample. The influence of storage conditions varies from one store to another and becomes more and more critical as time increases.
Table 2
Variance analysis of seed germination rate at one month of storage
Stores | Sample 2 | Sample 3 (1 month later) |
Fo_Boure | 61 a | 50.25 a |
SCDA_N'dali | 67.75 a | 60.25 a |
SCDA_Sinendé | 67.25 a | 52.50 a |
Sèkèrè | 51.75 b | 51 a |
Sokka | 62 a | 58 a |
Sonnoumon | 62 a | 51 a |
Yarra | 64.25 a | 51.50 a |
Moyenne | 62.29 | 53.5 |
Pr | 0.001** | 0.038* |
F | 5.831 | 2.78 |
CV (%) | 7.1 | 8.88 |
F: F of Fisher Pr: probability *: significant at 5%; **: significant at 1% CV (%): Coefficient of variation. On each line, the averages followed by the same alphabetical letter are not different at the 5% threshold according to the Student-Newman-Keuls test.
Figure 2: Boxplot of seed germination rate (Source: field data, 2019)
Comparative analysis of the three samplings with regard to estimated germination rate
Table 3 shows the variation of germination rates of the seeds according to the samples and the stores of origin. It appears in a global way that the seed germination rate varies very significantly with time and according to the different stores of origin (overall probability = 0). A variation ranging from 69.82 ± 2.33% to 52.7550 ± 5.11% is generally observed during the first two months of conservation, ie a loss of 17.07 ± 2.78% of germination power. A sharp decrease in the germination rate is observed at the third harvest, especially for seeds from the SCDA Sinendé and Yarra store, SCDA N'Dali and Sonnoumon. The stores are classified in two categories according to the Student Newman Keuls comparison test carried out after the third sampling. This is the category of stores with an average germination rate (Sonnoumon 50.75%, Sekere 50% and Fo_Boure 50.5%) and the category of stores with a good germination rate which consists of the four remaining stores. (SCDA_N'dali, SCDA_Sinendé, Yarra and Soka.
Table 3
Evolution of seed germination rate by time in different stores
Sources | Sample 1 | Sample 2 | Sample 3 | Probability |
Overall mean | 69.82 ± 2.33 a | 61.04 ± 5.65 b | 52.75 ± 5.11 c | ≈ 0*** |
Fo-Boure | 70.5 ± 1.29 a | 58.5 ± 2.08 b | 50 ± 5.60 c | 6.31.10-05*** |
SCDA N’Dali | 71.25 ± 0.96 a | 67.25 ± 2.50 b | 59.75 ± 0.50 b | 8.89.10-06*** |
SCDA Sinende | 69 ± 2.16 a | 65.5 ± 4.43 a | 51.5 ± 3.10 b | 1.01.10-04*** |
Sekere | 66.50 ± 1.73 a | 53 ± 4.24 b | 50 ± 2.16 c | 5.02.10-05*** |
Soka | 70.50 ± 1.29 a | 58.25 ± 3.59 b | 56.75 ± 2.06 b | 4.94.10-05*** |
Sonnoumon | 71.75 ± 0.96 a | 62.50 ± 2.38 b | 50.75 ± 1.50 c | 1.21.10-07*** |
Yarra | 69.25 ± 3.30 a | 62.25 ± 5.74 a | 50.50 ± 7.85 b | 4.84.10-03*** |
On each line, the averages followed by the same alphabetical letter are not different at the 5% threshold according to the Student-Newman-Keuls test.