The plant material is made up of axis of 420 individuals of different ages chosen randomly and arbitrarily in natural stands according to two types of habitats. The criteria for sorting the age categories, the dendrometric characteristics and the number of individuals per age category and per habitat type as well as the age types examined are recorded in Table 1.

According to the observations, the architectural development of the species is sympodial monochasial with only delayed relaying (delayed growth and branching). The terminal bud dies after a certain period of elongation, the axillary bud closest to the terminal bud takes over and so on. The plant appears as a pseudo-monopod with a twisted trunk structure. Parkia biglobosa can be classified in the architectural model of Paul Champagnat (1921-2011) and also of Wilhelm Troll (1897-1978) because of the way it builds its structure. Indeed, according to our remarks, the development of the species is a repeated complex of unbranched and discretely branched axes with short branches. The axes are vertical at their base and their distal end is by large radius curvature, reduced to horizontal or collapsed due to gravity and the flexibility of the wood. The leaves are spirally inserted and the flowering is sometimes lateral, sometimes terminal. In the young stage, the plant adopts the Champagnat pattern, while in adulthood and old age, the Troll pattern is more expressive.

Globally, the number of growth modules carried by the observed axes varied from 2 to 26 with an average of 10.08 ± 4.24; the average length of the Modules varied from 3.85 to 51.40 cm with an average of 18.57 ± 8.15 cm; the average diameter of the Modules varied from 0.43 to 8.81 cm with an average of 1.43 ± 0.82 cm. The average number of phytomers counted per Module varied from 5.5 to 33.5 with an average of 15.09 ± 5.35. The morphology and limits of the growth modules are presented in Figure 2.

Figure 3 shows the average distribution of the number and length of phytomers per growth module (MOD) as well as the growth stop or growth boundary zones (GS) along a Parkia biglobosa axis.

Analysis of the variance of modules morphological dimensions present on the observed axes according to the two environments considered (Table 3), showed that the morphology of the modules of the axes exposed to sunlight is statistically identical to the morphology of the modules of the axes benefiting from a superior shelter (P > 0.05). The environment in which this species evolves does not seem to have an effect on the dimensions of the modules. In Parkia biglobosa, the type of habitat (bright or dark light) does not significantly influence the morphology of the growth modules (P > 0.05).

Values with the same letters are not statistically different at the 5% level. Tree-H (m)= tree height in meters; Tree-Diam (cm)= diameter at breast or crown height of the tree in centimetres; Axis-Nbr-Mo= number of modules or growth units carried by the measured axis; Mo-leng (cm)= average length of modules or growth units in centimeters.; Mo-Diam (cm)= average diameter of modules or growth units in centimetres; Mo-Phyto-Nbr= average number of phytomers carried by the modules or growth units; Min= minimum et Max= maximum.

The comparison of the measured parameters (Table 4) according to the locations, showed a significant difference between the locations visited for each parameter of module morphology evaluated (P < 0.05). All module dimensions were neither increasing nor decreasing according to the arrangement of the localities along the defined South-North vegetation gradient. However, the modules were well developed in the localities of Bouaké and Katiola in the south of the defined gradient. In Parkia biglobosa, the morphology of the modules differed according to the characteristics of the study area (P < 0.05).

The results showed that the majority of the morphological parameters of the modules differed significantly between the ages (P < 0.05), except for the mean diameter of the modules, which remained statistically identical regardless of age (P > 0.05). The number of modules per axis was higher in mature trees, while the

Values with the same letters are not statistically different at the 5% level.

Values with the same letters are not statistically different at the 5% level.

average module length was higher in younger trees (Table 5). Old and mature trees had higher numbers of internodes per module than younger trees (Table 5). In Parkia biglobosa, the age of the individuals influenced the morphological dimensions of the growth modules (P < 0.05).

The average diameter of the growth modules did not vary significantly between seasons (Table 6); humidity did not have a major influence on the diameter of the growth modules (P > 0.05). In contrast, the number of modules per axis observed, the average length of growth modules and the average

Values with the same letters are not statistically different at the 5% level.

number of phytomer carried by growth modules differed significantly between seasons (P < 0.05). The majority of the dimensions increased during the rainy season (Table 6). In Parkia biglobosa, humidity influenced the majority of the morphological dimensions of the growth modules (P < 0.05), except the mean diameter (P > 0.05).

The results showed that the boundaries of the Modules are represented by successively very short internodes (Figure 2 and Figure 3). These observations were found on Guayule (Parthenium argentatum) [37] [38] [39] and on Parkia velutina [1] etc. Parkia biglobosa has a sympodial architectural development with a single relaying axis (monochasial sympod), in contrast to Guayule (Parthenium argentatum) which is a sympod with several relay axes (polychasial sympod). Guayule belongs to the architectural model of Anthony Leeuwenberg like cassava. [34] indicated that Parkia biglobosa, of the Fabaceae family had an architectural development following the model of Wilhelm Troll, we think that the architectural model of this species also integrates that of Paul Champagnat according to our observations. This remains to be verified in further studies on the architectural analysis of Parkia biglobosa.

In our study, the environment in which the individuals evolved did not significantly influence the morphological dimensions of the growth modules (P > 0.05). This is due to the fact that we considered individuals of all age types observed in this environment in a global manner. The finding might be different if the observations on morphological dimensions were made only and separately by age category. Alternatively, this fact is due to the nature of the species; indeed, some species respect their own genetic programme of architectural development regardless of environmental stress [34]. Moreover, this species is a slow-growing savannah [9], and therefore the immediate perception of the effect of light on shoot growth might be difficult. Normally in a given plant, individuals in the shade (undergrowth) should be less developed than those exposed to the sun (full sun) as in the studies of [3] on Fagus sylvatica and of [40]. However, plants growing in the shade are very often tall and etiolated because of the terminal bud’s search for bright light for photosynthetic activity. Indeed, bright light is the key factor for growth, it guarantees a good growth in length and width of plants contrary to dark light [40] [41] [42].

The variation in morphological parameters between locations is due to the specific and distinct environmental characteristics of each location surveyed (climate, soil type, rainfall, relative humidity etc.). We had thought that the morphological dimensions of the measured Modules would be ordered in an increasing or decreasing manner according to the layout of the surveyed localities. This was not the case; in fact, even if the surveyed localities are arranged according to a “pseudo-climatic” gradient of dense vegetation decreasing towards the north, each locality has its own distinctive characteristics. Moreover, in the current context of climate change, climatic and microclimatic parameters are very unstable, uncontrolled and vary greatly within a day. Several studies have shown that the depth and fertility of the soil, the environment and even the age of individuals influence the architectural development of species [30] [42] [43]. [44] and [45] have also indicated in their research that climate or an ecological gradient has an effect on plant morphology. Similarly, studies by [46] and [47] [48] indicated that the origin of differences in tree morphology is due to factors such as soil type and the genetic characteristics of the individuals sampled. On the other hand, the modules were well developed in the southern part of the defining gradient (Bouaké and Katiola). This area could be a favourable and priority area for establishing permanent plots and increasing the success rate of reforestation or agroforestry programmes using this species.

The morphological dimensions of the modules evaluated varied with the age of the individuals (P < 0.05). The number of phytomeres per module is higher in old and mature trees, due to their age and the preformation of organs in the latent buds. In young trees, module length is higher due to the growth dynamics of young shoots (more intense meristematic activity). Indeed, some studies have shown that the age of individuals influences tree morphology [31] [43] [47] [48]. Age strongly modifies the morphological dimensions of plant organs regardless of location, due to very different functional capacities (number of photosynthetic leaves, evapotranspiration capacity, mineral storage, root suctions etc.) and structural dimensions (root and leaf sizes, space occupancy etc.). Young individuals with well-developed modules than the other two categories could be suitable ideotypes in the search for resilient genotypes adapted to global change for the establishment of plots based on this species by cuttings.

In the rainy season the size of the modules increases. This period is favourable to the synchronised bud break of the terminal buds of the axes. In French Guiana, [1] observed the same phenomenon on Parkia velutina. The meristems of the buds, which have long remained dormant or in slow growth due to a long dry season, take advantage of this period to maximise their development. Indeed, water is vital for plants and its availability guarantees a good physiological and harmonious functioning of plants (metabolism, photosynthesis, organogenesis etc.). Whereas water stress slows down all physiological activity and causes their decline. [49] has indicated in his research the positive effects of water regime on plant growth and development. In the process of establishing a plot or implementing reforestation or agroforestry programmes based on Parkia biglobosa, the rainy season remains the best and most favourable period.