Evaluation of the Agronomic Performance of Onion Varieties in the Rainy and Dry Season in Northern Ivory Coast

Onion is a vegetable of great consumption in Côte d'Ivoire. However, it is only grown in the dry season and national production cannot meet the needs of the population. The objective of this study was to identify onion varieties adapted for production during the rainy season in northern Côte d'Ivoire. To better assess the performance of the varieties, trials were conducted in both the dry and rainy seasons. Six varieties were evaluated, including : Galmi Violet, Damani Violet, Safari, Karibou, Ares and Syng 8362. Measurements included plant survival rate after transplanting, growth parameters, flowering rate, production parameters and bulb quality parameters. In the dry season survival rates ranged from 95.64% to 97.56%. The yield of Galmi Violet was the highest (27.09 t/ha) and the lowest was noted at Safari (16.21 t/ha). In the rainy season, survival rates varied between 37.12% and 88.58% and yields varied between 3.19 t/ha and 23.15 t/ha. With the exception of Ares, all other varieties lost half their density. Ares was the most efficient with a total yield of 23.15 t/ha.


INTRODUCTION
Onion (Allium cepa L.) is a herbaceous monocotyledon belonging to the family Amaryllidaceae (APG III, 2009). It is cultivated as a vegetable plant for its bulbs and leaves, which are basic ingredients for many culinary preparations in all regions of the world (Megroz & Baumgartner, 2000). In sub-Saharan Africa, onion cultivation is practiced in all countries and the sector is an important source of income for all those who engage in it (Cathala et al., 2003). However, except Niger and Burkina Faso, which produce quantities that exceed domestic demand, the production of other West African countries cannot meet national consumption needs (Tarchiani, 2013).
In Ivory Coast, national onion production is between 5.000 and 7.500 tons/year, while the annual demand for onions is now estimated at over 100.000 tons. Due to this large difference between supply and demand, Ivory Coast is dependent on the external onion market for about 95% (Anonymous 1, 2014). In order to reduce Ivorian onion imports, SODEFEL (ex-Société de Développement pour la production de Fruits et Légumes) had initiated cultivation in the north of Ivory Coast in 1980. This action allowed the selection and extension of the Galmi violet, a Nigerian variety introduced in Côte d'Ivoire (Fondio et al., 2001). In the same vein, Silué et al. (2003) undertook to select other onion varieties adapted to northern Ivory Coast. Despite these actions and the constant supervision of producers by the National Rural Development Support Agency (ANADER) to increase onion production, it remains low. This low onion production in Ivory Coast is due to climatic conditions that limit cultivation essentially in the north (Silué et al., 2003). Moreover, it can be explained by the fact that onion production is almost entirely in the dry season (Ali et al., 2018). To increase Ivorian onion production, one possibility would be to extend the production period in the rainy season with varieties that are adapted to it. The objective of this study is to identify highyielding onion varieties adapted for production during the rainy season in northern Ivory Coast.

MATERIEL ET METHODES Study site and vegetal material
The experiments were carried out on the plot of land of the National Rural Development Support Agency (ANADER), located in Ferkessédougou (9°62'63 north latitude and 5°32'72 west longitude), in the north of Côte d'Ivoire. The department of Ferkessédougou is subject to a monomodal rainfall regime, covering the period from May to October and a dry season from November to April. The dry season is characterized between December and February by a cool and dry wind, the harmattan. The climatic data (rainfall, relative humidity, minimum and maximum temperature, exposure time and photoperiod) recorded during the experiments are shown in Table 1. The soil texture is predominantly sandy with a pH of 6.5.

Conduct of trials and agronomic practices
The rainy season test was conducted from June 2018 to November 2018 and the dry season test was conducted from November 2018 to April 2019. The production method used was transplanting survival nursery seeding. The nurseries were set up in beds of 2 m 2 in area according to a row seeding. The 45-day old plants were then transplanted into 6 m 2 elementary plots in a fully randomized Fisher block system with four replicates. The transplanting geometry was 15 cm between rows and 10 cm between plants on a single row. Planting density was 50 plants/m 2 , or 500.000 plants/ha. Before nursery sowing and transplanting, the beds were disinfected with Thioral, a mixture of insecticide and fungicide. Fertilizer NPK 18-24-24 was applied in the nursery (10 g/m²) and before transplanting (20 g/m²) as a background fertilizer. Urea 46% was applied in the nursery (5 g/m²) 25 days after sowing and in the field (20 g/m²) one month after transplanting. A second application of NPK 18-24-24 (20 g/m²) was made in the field, one month after the urea application. The frequency of watering in the field was three times a week in the dry season and on demand in the rainy season.

Quality parameters
The dry matter content of the bulbs was determined by differential weighing in accordance with French standard NF V 03-707 (2000) on 5 bulbs per variety. To do this, 10 g of onion slices were weighed before and after being placed in an oven at 100 °C for 24 hours. The dry matter content was deducted by difference from the moisture content and then expressed as a percentage. The dry soluble extract expressed in degree Brix (° Brix) was also determined on five bulbs per variety with a digital refractometer type PR-1 Atago.

Flowering rate and length of crop cycle
For each variety, the flowering rate was obtained by dividing the number of flowering plants by the number of living plants, multiplied by 100. The length of the cropping cycle considered in this study is that from sowing to harvesting of the bulbs Statistical analysis All data collected were subjected to an analysis of variance (ANOVA) using STATISTICA 7.1 software. If there were significant differences between the means, the Newman-keuls multiple comparison test at the 5% threshold was used to classify them. The data expressed in percentages were transformed according to the formula arcsin√x/100 before being submitted for analysis.

RESULTATS ET DISCUSSION
Results of the variety effect on plant establishment, growth parameters and flowering rate under both climatic seasons are presented in Table 2. In the dry season, plant establishment percentage ranged from 95.64 % to 97.56 %. The differences observed were not significant (P > 0.05). Plant mortality was low in all varieties, which could be explained by their good adaptation to the environmental conditions of the dry season. Sensitive varieties would have been strongly affected. On the other hand, in rainfed crops, plant survival rates after transplanting ranged from 37.12 % (Safari) to 88.58 % (Ares). Analysis of variance showed a very significant varietal effect (P < 0.001). The varieties Violet of Galmi, Karibou, Safari, Violet of Damani and Syng 8362 recorded losses of more than half their density. Ali et al. (2018) obtained similar results when searching for onion varieties adapted for rainfed cultivation in central Burkina Faso. These varieties are reported to be very sensitive to excess moisture. Therefore, they do not seem to be adapted to rainfed cultivation. The survival rate of 88.58% for the Ares variety attests to its suitability for winter cultivation in Ivory Coast. This result confirms that of Anonyme (2014) obtained in Ghana.
As far as growth parameters are concerned, they were all influenced by the variety regardless of the climatic season (P < 0.001). In both the dry and rainy seasons, Ares was the highest with heights of 50.55 cm and 51.63 cm respectively. Safari was the shortest (40.30 cm in the dry season and 42.87 cm in the rainy season). In the dry season, the number of leaves issued by Ares (8.6), Karibou (8.4) and Violet de Galmi (8) were statistically equal and higher than those of the other varieties. The lowest numbers were noted in Syng 8362 (6.13) and Safari (6.53). In the rainy season, Ares produced more leaves (11.4) than the other varieties. Comparable results were obtained by Ahmed et al. (2013), who attributed the observed differences to genetic factors of the varieties used. Moreover, for each variety, the values of height and number of leaves emitted recorded in rainy season cultivation are relatively higher than those obtained in the dry season. These differences are thought to be related to environmental factors, including rainfall and temperature (Tesfay et al., 2011). In this regard, studies by Kumar et al. (2007) showed that onion growth parameters such as plant height, number of leaves and plant biomass have a positive correlation with the amount of water received. Also, according to Bosekeng (2012), the relative growth rate of leaves increases linearly with a temperature of up to 20°C. Beyond this threshold, the plant growth rate begins to decrease and when the temperature exceeds 26°C, growth ceases completely.
Bulbing index fluctuated between 5.47 (Violet of Galmi) and 4.08 (Safari) in the dry season and between 4.54 (Ares) and 3.16 (Safari) in the rainy season. In both the dry and wet seasons, the bulbing index (BI) of all varieties was above 2, the reference value of (Clark & Heath, 1962), which means that they all bulbified. However, with the exception of Ares, the bulbing index of the other varieties showed a major reduction during rainfed cultivation. This could be related to a reduction in photosynthetic activity due to low sunlight and high relative humidity. Indeed, the bulb is the organ of reserve of the photosynthetic assimilates. Under the effect of high atmospheric water vapour, the rate of photosynthesis and water uptake by plant roots is reduced due to partial or complete closure of stomata (Brewster, 2008). The good suitability of Ares is based on the fact that the performance of a variety depends on the interaction between its genetic factors and environmental conditions (Jilani & Ghafoor, 2003).
The flowering rate was also influenced by the variety but also by the season. The varieties flowering in summer cultivation are Ares (3.12 %), Safari (1.88 %), and Violet of Galmi (1.34 %). In the rainy season, only Ares (1.72 %) flowers. Onion flowering occurs at the end of the vegetative period following a period of vernalization (Aoba, 1960). Thus, cold stimulates the accumulation of cytokinin and gibberelin, altering the hormonal balance and leading the plant to develop a flowering stem (Rakhimbaev & Ol'Shaskaya, 1976). The absence of flowering observed in some varieties in both climatic seasons could mean that the length of vernalization required to stimulate flowering varies between varieties. Furthermore, Rouamba (1993) mentions that temperature is not the only factor inducing flowering in onions. He reports a genetic origin to explain the first year flowering observed in some onion varieties.  Table 3 presents the length of the cropping cycle of the onion varieties studied and the results for the production components during the two climatic seasons. The effect of the variety was significant on all parameters. The crop cycle of Syng 8362 was 145 days in the dry season compared to 150 days in the wet season. The crop cycle of the other varieties was 152 days in the dry season and 158 days in the wet season. This shows that Syng 8362 is the earliest variety. The differences in days observed between the crop cycles in the dry and rainy seasons are thought to be caused by the rains. According to Doorenbos and Kassam (1979), excessive irrigation during bulbing-out delays bulb ripening. Consequently, it prolongs the crop cycle.
The rate of bulb rot recorded at harvest differs between varieties. In the dry season, rot rates ranged from 0% for Ares to 1.24% for Syng 8362. In the wet season, values ranged from 0.72% for Ares to 5.20% for Syng 8362. Bulb rot is caused by excess water (Collin et al., 2004), which explains the increase in rates observed in the rainy season. The Syng 8382 variety that recorded more rot would be more susceptible to excess moisture. Total bulb yields ranged from 16.21 t/ha to 27.09 t/ha in the dry season and from 3.19 t/ha to 23.15 t/ha in the wet season. The differences observed were very significant (P ˂ 0.001). In the dry season, the highest gross yields were recorded for Galmi Violet and the lowest were recorded for Safari. The low yield recorded at Safari can be explained by its low productivity. In the rainy season, with the exception of Ares (23.15 t/ha), yields of the other varieties were very poor. They could not express their dry season performance. This can only be explained by the unsuitability of these varieties for rainfed cultivation.
In the dry season, the average bulb weights of Violet of Galmi (80.53 g), Syng 8362 (76.69 g) and Ares (76.35 g) were statistically equal and higher than those of the other varieties. The lowest value was noted in Safari (59.14 g). The majority of bulbs produced by Violet of Galmi (64.19%), Ares (59.91%), Syng 8362 (54.89%) and Karibou (54.38%) were of size C3 and those of Safari (56.49%) and Violet of Damani (51.64%) were of size C2. In rainfed cultivation, average bulb weights ranged from 24.50 g to 70.52 g. The average weight of Ares bulbs was the highest and the majority of its bulbs (57.38%) were C3 size. On the other hand, the majority of the bulbs produced by the other varieties were of size C2. The average bulb weights of the latter were much lower than those obtained in the dry season. The predominance of small-sized bulbs confirms the generally poor performance of these varieties in rainfed cultivation and explains their low yields.
After subtracting rotten bulbs and C1 size bulbs from the gross yield, marketable yields fluctuated between 15.11 t/ha (Safari) and 26.27 t/ha (Violet of Galmi) in the dry season and between 2.59 t/ha (Violet of Damani) and 22.43 t/ha (Ares) in the rainy season.  The results obtained corroborate those of Abhayawick et al. (2002) who reported that the dry matter content of onion bulbs normally varies between 7% and 18% of the fresh weight depending on the variety. This is a parameter that also deserves special attention from the point of view of storability. Indeed, there is a positive correlation between the general aptitude for conservation of the onion and the dry matter content of the bulbs (Assogba et al., 2007). A variety with a high dry matter content is a variety with a better bulb keeping quality. Onions with high dry matter content are firmer and therefore more resistant to damage caused by transport and handling (Silué et al., 2003). For each parameter and per season, the means assigned the same letter are statistically identical to the 5% threshold (Test of Newman-Keuls).

CONCLUSION
The results obtained from the trials showed that the performance of onion varieties differs according to the climatic seasons. In the dry season, Galmi's Violet variety performed best, surviving Syng 8362, Ares, Karibou. Safari and Damani Violet performed less well. In the rainy season, Ares was the only performer. Other varieties including Violet of Galmi performed less well. Thus, Ares is the best variety to grow for rainfed production in the north of Côte d'Ivoire. Other onion varieties suitable for rainfed cultivation must be identified in order to make several varieties available to producers.