The studies were carried out at the Centre National de Floristique (CNF), located within the Université Félix HOUPHOUËT-BOIGNY in Abidjan. It is located at longitudes 5˚20'51" North and latitudes 3˚59'01" West. The climate at the CNF experimental site is of the Atetian type, characterized by four seasons determined by rainfall [11] . The soil is essentially ferralitic [12] . The soil type is characterized by the presence of a thin humus horizon and a thin gravel horizon. The study area is home to most of the tree species of the dense evergreen rainforest. They are carefully preserved in the CNF botanical garden.

The plant material used in this study consisted of Desmodium adscendens cuttings with four buds. They were taken from three-month-old plants grown in nurseries. They were taken from nursery plants at the Société Agricole Kablan Joubin (SAKJ) site in the village of Akréssi in Ayamé department.

The shade structure was covered with layers of shade to attenuate solar radiation to the desired level and protect the plants from insect attack. These shelters were 5.29 m long and 3.10 m wide. The level of shading was determined using a luxmeter placed above and below the layer of mosquito netting. The measurement consisted of determining the amount of light able to pass through one, two and three layers of mosquito netting. A total of four shading levels were selected. Level 1 (N0) corresponds to a shelter with no shading, and is the control treatment. Levels N1, N2 and N3 correspond to 20%, 40% and 60% light, respectively.

Three-month-old Desmodium adscendens mother plants from the nursery were selected on the basis of their physiological and health status [13] . Cuttings were obtained by cutting into 5 cm to 10 cm portions of the stem using pruning shears below a node. Cuttings with four nodes and leaves were taken [14] and underwent partial leaf removal [15] . Transplanting was carried out without removing the leaves inserted at the nodes to avoid damaging the buds in the leaf axils. This transplanting method has been described and used for cuttings of many ornamental species. The soil was taken from the National Centre de Floristique. The 10-litre trays were filled with soil previously sterilized in an oven set at 120˚C for one hour. Holes were drilled in the substrate and the cuttings were inserted vertically below the third node. After planting, light pressure was applied around the cuttings.

The trial was carried out to study the effect of shade levels on the growth and development of Desmodium adscendens seedlings ( Figure 1 ). The experimental set-up consisted of four defined shelters representing the four treatments: control N0 (no shading), N1 (20% shading), N2 (40% shading) and N3 (60% shading). For each treatment, ten trays were used in which five cuttings were planted, for a total of 50 plants per treatment. All production factors, i.e. substrate, shade level, irrigation and manual weeding, were kept uniform and at their optimum for each treatment. Each tray was watered every other day with 1 liter of water.

The recovery time is the time elapsed between the planting of the cutting and the appearance of the first bud. The vegetative recovery rate (Tr) was obtained by dividing the number of cuttings that recovered, i.e., those that produced buds (Nr), by the number of cuttings initially transplanted (Nt). The formula is as follows:

Tr (%) = Nr/Nt × 100 (1)

The maximum recovery rate corresponds to the cutting recovery capacity of the species. The recovery rate is the variation in the recovery rate of cuttings as a function of time. The ratio between the cumulative number of cuttings that have given buds daily and the number of cuttings planted.

The cuttings mortality rate (Tm) was obtained by dividing the number of dead cuttings (Pm) by the number of cuttings initially transplanted (Nt), and multiplying by 100 [16] . In our case, this parameter was taken into account at the end of the trial and determined using the following formula:

Tm (%) = Pm/Nt *100 (2)

Stem height (H) was the distance from the collar to the apical apex. It was measured every week for 8 weeks using a 50 cm-long graduated tape measure. These values were used to determine the evolution of cuttings height over time. The diameter of the plant stems was the volume or thickness of the stems. Measurements were taken every week using a caliper graduated to the millimeter. In the case of overestimation of volume, the diameter reading was given without rounding down or up the values obtained.

During the trial, the number of leaves (F) emitted per plant was counted every five days for 75 days. These measurements made it possible to evaluate the evolution of the number of leaves as a function of time. Spread was determined on the basis of the number of branches per plant. It was assessed every week for 3 months.

Flowering rate is the monthly variation in the number of flowers as a function of time. The number of pods was counted every three days. The average number of pods per plant was determined by dividing the number of pods by the number of plants assessed. These values were used to determine changes in the number of pods per plant as a function of time.

The mass of pods obtained for each treatment was measured using a balance. The seeds were then oven-dried at 50˚C, and regular measurements were taken to determine the dry mass. The water content of each pod was determined using the following formula:

Water level (%) = (mf − ms)/mf × 100 (3)

After 3 months, the plants were uprooted and transported to the UP Plant Physiology and Pathology Laboratory. Fresh mass (Mf) was assessed by weighing the roots, stems and leaves, which were then placed in a 50˚C oven. Regular weighing determined the dry matter (Ms). The dry matter content of each organ was determined using the following formula:

Qs (%) = Ms/Mf × 100 (4)

The length of the Desmodium adscendens growing cycle was determined from the date of transplanting of the cuttings to harvesting. It takes into account the phrenological phase of vegetative growth, flowering and fruiting. The vegetative growth phase begins as soon as the cuttings are transplanted and ends with flowering. The flowering phase was taken into account when the flower buds set and fruiting ended when the fruits ripened.

The data obtained were analyzed using STATISTICA version 7.1 software. An analysis of variance (ANOVA) was used to study the effect of different shading levels. In the event of a significant difference between the means, the Newman-Keuls comparison test at the 5% threshold was used to separate the means into homogeneous groups.

The lag times were 2 days and recovery began on the 5^th day and continued until the 11^th day. The curves for treatments N1 and N2 showed an increasing trend. These curves were well above the other curves, fluctuating between 22% and 78% from day 3 to day 11. The N3 curve evolved intermediately between 16% and 64% from D3 to D11. The N0 curve fluctuated between 6% and 12% and evolved below the other curves. The analysis revealed significant differences between treatments. The cuttings recovery rate varied from 52% to 78% ( Figure 2 ). Treatment N0 had the lowest recovery rate at 52%. Treatments N1 and N2 had high recovery rates of 74% and 78% respectively. Finally, intermediate values were observed for shade level N3 (66%). Analysis of variance showed significant differences (F = 6.096; p = 0.002) between mortality rates for the different treatments ( Figure 3 ). The lowest mortality rates were observed in treatments N1 (28%) and N2 (26%). On the other hand, high mortality rates were observed in the control (N0) at around 56%. Treatment N3 had 56% dead cuttings ( Figure 4 ). The effects of shade levels showed highly significant differences (F = 3.780; p = 0.012) between mean plant heights from week S4 recorded in Table 1 . Comparison of the means showed that the mean height of the plants under shade was higher than that of the plants exposed without shade. Treatments N1, N2 and N3 showed the greatest growth. The values were 3.43 cm and 4.18 cm at S1 before rising to 56.23 cm and 49.04 cm at S12. The unshaded N0 plants, on the other hand, grew more slowly, recording heights of 3.419 cm at S1 and 21.568 cm at S12. The average height of plants in treatment N3 increased progressively to reach a value of 40.333 cm in week 12. Analysis of the data revealed a significant difference (F = 4.884; p = 0.003) between plant thickness for the different treatments. Shade level N2 (40%) gave the best plant diameter for Desmodium adscendens and the average plant diameter was 1.859 cm. On the other hand, small diameters of Desmodium adscendens plants were obtained for N0 and N3 with diameters of 1.48 and 1.59 cm respectively. Treatment N1 had intermediate Desmodium adscendens plant diameter values of 1.62 cm ( Table 2 ). The analysis revealed a variation in the average number of leaves between treatments. Considerable changes were noted for treatments N1 and N2. The number of leaves varied from 0.58 to 85.48 between D0 and D75. With N3, the number of leaves ranged from 0.46 to 38.32. The lowest number of leaves was noted for treatment N0 with only 28 leaves at D75 ( Table 3 ). The branching curves showed exponential and continuous growth. The first branches were observed from day 5. The N1 and N2 curves were well above the others, while the N0 curve was below. The N3 curve evolved intermediately ( Figure 5 ). After 9

weeks, analysis of the data showed a significant difference. Plant branching was greatest in treatments N2 (11% or 22%) and N1 (10% or 20%), followed by treatments N3 (8% or 16%). Treatment N0 (6% or 12%) showed low branching, with a total of 6.95 ( Table 4 and Figure 6 ).

Analysis of variance showed significant differences (F = 3.041; p = 0.0347) between the number of flowers in the different treatments from day 72 onwards. The rate of flower emergence is shown in Figure 7 . The curves for N1 and N2 reached their optimum at 101 days. The gradual evolution of the number of flowers reached a peak at day 132 for N0 and N3. The analysis revealed significant differences between treatments. Plants under shade N1, N2 and N3 had the highest number of flowers, with values ranging from 14.8 to 17.4 flowers ( Table 5 ). The curves show three parts for all treatments. Between days 175 and 165, the N1 and N2 curves moved well above the other curves. The N0 and N3 curves showed little growth from day 105. Pod production was statistically identical for the treatments between days 45 and 75, while differences were observed between days 105 and 135. The number of pods was greater for treatments N1 and N2, with 39.93 and 36.76 pods on day 135 ( Figure 8 and Table 6 ). Analysis of variance revealed no significant difference (F = 1.131; p = 0.366) between treatments. Total fresh weight was greater in N1 (20.58 g) and N2 (24.26 g) than in N0 (7.6g). The fresh weight of the above-ground part was lower. There was no difference in plant mass or dry matter content between treatments ( Figures 9-11 ). The effect of shade levels on pod and seed weight showed a significant effect (p = 0.000). The fresh weight of pods per plant was low for the N0 treatments (0.128 g) followed by the N3 (0.55 g). On the other hand, the highest weights were noted for N1 and N2 with 0.96 g and 1.09 g respectively. For dry weight, the highest values were obtained for treatments N1, N2 and N3 with 0.38 g, 0.41 g and 0.27 g, respectively. As for the weight of dry seeds per pod, the results showed that N1 (3.29%) and N2 (4.99%) had the highest values ( Table 7 ).

Figure 12 shows the growing cycle of Desmodium adscendens as a function of shade levels. The length of the growing cycle is not the same for all shade levels.

In treatment N0, from transplanting of the cuttings to flowering, the cycle lasted 103 days, while in treatments N1, N2 and N3, it varied from 62 to 66 days. The lag time and leaf emission were 2 days for treatments N1, N2 and N3. Spreading of the N1 (20%) shade plants began at 30 days, the N2 (40%) at 34 days, followed by the N3 (60%) at 35 days, and finally the N0 plants at 46 days. The duration of the phase from flowering to harvest was 19 days for the N2 and N3 shading levels, and that of the N1 and N0 treatments was 13 and 17 days, respectively. The ripening time for pods produced in N0 was 9 days. It was 13 days for treatment N1, 15 days for treatment N2 and 18 days for treatment N3. Overall, the crop cycle lasted 116 days for treatment N0. On the other hand, for treatments N1, N2 and N3, it lasted practically the same number of days.