The Architectural Unit Setting up and Architectural Characteristics of Néré, Parkia biglobosa, Jack, R. Br. (Fabaceae)

Parkia biglobosa is a much-loved and over-exploited African savannah species for its socio-economic importance. Knowing and taking into account its architectural unit, which is the basis for diagnosing phenology, productivity and tree health, could provide a new perspective on its sustainable management. The aim of this study is to establish the architectural development in Parkia biglobosa by retrospective analysis. To achieve this objective, 390 individuals of all sizes ranging from seedlings to senescent trees were observed and analysed under various soil and climatic conditions in Côte d’Ivoire. The results showed that Parkia biglobosa is a light plant but shading tolerant. It is a mixed vegetative axis plant, the stem phytomere, the module or growth unit, the axis, the architectural unit and the reiterated complex. Retrospective analysis of the modules showed that the dimensions of the growth units are indicators of morphological variation and species adaptation to a changing climate (P < 0.05). However, the equations generated by the morphological and habitat dimension linkage models are not significant (R 2 and r < 0.7) to be used as a guide for field data collection. This study represents an initiation into the architectural study of this species and the information provided will serve as a basis for further research into the architecture in relation to the sustainable use of this species.

Parkia biglobosa by retrospective analysis. To achieve this objective, 390 individuals of all sizes ranging from seedlings to senescent trees were observed and analysed under various soil and climatic conditions in Côte d'Ivoire. The results showed that Parkia biglobosa is a light plant but shading tolerant. It is a mixed vegetative axis plant, the stem is orthotropic* in its proximal part and plagiotropic* (collapsing) in its distal part in young stage. The tree then transitions to an adult and old stage into a tree with a plagiotropic* axis in the proximal and distal parts, the trunk is built up by superimposing collapsed relay axes that gradually straighten, branching is sympodial*, growth is defined and sexuality is terminal and lateral. The ontogeny takes place in three phases: initiation of development and establishment of the crown (young), then flowering and establishment of the architectural unity (adult) and finally the death of secondary axes in the crown, duplication of the architecture by a series of partial and total reiterations (old). The level of organisation is 5: the *In this document, terms marked with an asterisk (*) are defined in a glossary appended to the article after the references in order to alleviate difficulties of understanding. Indeed, the vocabulary used differs from that commonly used by foresters.

Introduction
The architectural* study of a plant is a morphological approach that makes it possible to characterise the organisation of trees or groups of trees growing in different pedoclimatic contexts [1] [2] [3] [4]. In spite of this pedoclimatic variability, it nevertheless allows us to highlight a specific average architectural organisation that constitutes the species' sketch or architectural unit*. This sketch is the architectural expression of the plant that allows it to be visually differentiated from another tree species without having studied it, and in the second instance constitutes the pillar of its sustainable management in case of threat [5] [6] [7]. Indeed, the architectural unit constitutes the smallest stem structure necessary and sufficient for the plant to reach its sexual maturity phase (complete reproduction) and thus complete its life cycle by forming flowers and fruits [8] [9]. The plant then continues its growth by altering this architectural unit through the phenomenon of duplication; a series of axes of different categories are then reproduced in the adult plant [10] [11] [12]. It is, therefore, necessary and important to bring out this architectural unit hidden in the tree, as it is the basis for future diagnostics by foresters or observers to assess the difficulties, phenology, health, productivity and reproducibility of species in various geographical situations under stress or climate change.
Plant architecture* can be achieved through retrospective analysis, which analyses the structure of the plant using various morphological markers that allow the formation and successive increases in length of the axes to be limited and dated. It depends on the spatial and temporal arrangement of the plant parts (branched hierarchical system), and is based on these morphological traits at the shoot and branch level (axes grouped into categories characterised by morphological criteria, exploration and/or reproduction role); then affected by endogenous (genetic) and exogenous (environmental) factors [10] [13]. The growth pattern through which the plant develops its form is the architectural model* of the plant or the basic growth strategy of a plant. The main architectural parameters generally studied are growth, branching, morphological differentiation of  [11]. This discipline has long been recognised as an important scientific tool in horticultural crops, understanding plant function, yield assessment and for the development of crop models [14] [15] [16]. It has been used extensively and successfully in Europe and South America for yield optimisation, forest management, preservation of important species, understanding the adaptation of species to climate, sustainable exploitation and safeguarding of threatened species [1] [2] [6] [7] [17]. However, it has never been applied or applied to emblematic West African species, yet this area is a strategic carbon sink encompassing many important overexploited and threatened species. This is the case of Parkia biglobosa, an agroforestry species indigenous to the savannahs of Africa. This species is much loved by the rural populations of this geographical area for its socio-economic role [18] [19] [20]. As a result, it is one of the most overexploited species among many others, but also the best documented to date [19] [21]. However, its architectural study does not exist in the literature to our knowledge. However, the knowledge of its architectural development sequence, its architectural characteristics and its architectural unity are very important for its phenological diagnosis, the evaluation of its productivity and its preservation. Indeed, in the first step, the architecture of this species completes its biological knowledge and in the second step, it allows access to the structure-function-time-environment relationship and thus can give a new point of view on the sustainable management of this species. The objective of this study is to establish the architectural development in Parkia biglobosa via retrospective analysis of various individuals from different environmental conditions in Côte d'Ivoire.

Trees Studied
390 freely growing individuals in open and forest environments were arbitrarily selected and observed. These individuals of different ages (young, adult and old) were derived from wild individuals of natural regeneration (forest) and artificial regeneration (1-and 2-year-old individuals planted and monitored in a nursery) benefiting from a canopy of variable size. The number of individuals as well as the dendrometric characteristics per age category and environment, and the location of the growth units or modules studied are recorded in Table 1.

Study Sites
The study was carried out in seven locations along a bioecological gradient in Côte d'Ivoire ( Figure 1). The soil and climate characteristics of the surveyed locations are shown in Table 2

Retrospective Analysis: Choice of Axis Type and Habitat
Observations were made on different types of axes depending on the accessibility of the crowns in two different habitats. For young trees, the axes assessed were the main trunks because of the easy access and the non-frequency of secondary branches on all individuals in this category. For mature and old trees, the axes assessed were only tertiary branches and short twigs due to accessibility. Assessments were carried out in situ for young trees; whereas, for mature and old

Parameters Assessed
To qualify (architectural analysis) and quantify (retrospective analysis) the axes in the architectural approach and to facilitate the drawings and subsequent description of the architectural characterisation, the following characteristics were (mature and old trees).

Statistical Data Analysis
Quantitative data were pooled and compared with each other (MANOVA) using SAS software version 9.4. The Student-Newman-Keuls test at the 5% threshold was used for post hoc comparisons. The links between the different quantitative parameters were made using XLSTAT 2020 version 7.5.

Results
Through this work, which required the observation of several individuals of all sizes (from seedlings to senescent trees) by means of synthetic drawings, the de-

Young Stage
In Parkia biglobosa, the seedling consists of an unbranched stem with bipinnate compound leaves. The phyllotaxy is alternate spiral. At this stage the leaf bears 2 to 6 leaflets with one stipule at the base of the petiole (leaf sheath) and another stipule at the end of the primary rachis. Each leaflet bears 9 to 16 pairs of secondary leaflets and the secondary rachis bear a single secondary stipule inserted at the end of the secondary rachis. The young stem tends to zig-zag (right-leftright-left) in the direction of the weight of the established leaves. After the formation of a phytomere, the stem continues to grow in the opposite direction to the previous leaf. This causes the main stem to twist and gives the impression that the leaves are distributed in an alternating spiral fashion along the leaf stem, but in reality the phyllotaxy is alternating distichous. After 3 to 4 months of evolution, the apical meristem dies by desiccation, the part below the dead part swells by bulging, giving a water tower shape to the phytomere. At this point, the initially monopodial growth mode becomes sympodial. In the forest, nodule-like swellings appear on the leaf rachis and the main stem at this stage. The nearest axillary bud takes over in the direction of the main stem and after 2 to 6 phytomeres, it dies in turn, and so on. This system forms clearly visible modules along the stem. In most cases there is only one relay shoot and the structure is a monochasial* sympod. The stem is usually curved at the distal end and later straightens as the stem expands and loses its leaves. In fact, as the stem grows in height, the old leaves (in the base) are pruned off by themselves and in succession. The stem retains only the new leaves of the new module in the apical part. These leaves are larger and heavier than the phytomer, the module and often the stem; it is the weight of the leaves that causes the stem to collapse. After a year, when the main stem bends, the relay bud sets up in the bending part of the stem, the rest of the bent portion prunes and so on. The stem is a pseudomonopod* and the modules are clearly visible on the stem (growth arrest zones are marked by markers and well-swollen plateau-like areas). The branching appears in the second year, the branches are short, deferred* and located in the middle of the tree. The branching is initially in a vertical direction and collapses, and then lengthens in the horizontal direction. It forms in the middle of most modules (mesotone*) and often at the end of the modules (acrotone*). The branching is not continuous, it does not follow any growth rhythm, it is diffuse* on the trunk. All the branches formed follow the same process of pruning the trunk and are all sympodes. This is the beginning of the establishment of the top. Observations on individuals of this stage revealed that the individuals cultivated and monitored in the nursery were larger and more vigorous than the wild individuals observed in the forest. The latter suffer more trauma (insect attacks) than those observed in nurseries. Table 3 shows the comparison of the morphological parameters of the modules or GUs according to the habitat, the localities surveyed and according to the habitat per locality. The habitat does not statistically influence the morphology of the modules (P > 0.05). However, most of the morphological parameters vary from one locality to another (P < 0.05) and from one habitat to another per locality (P < 0.05).

Mature Tree and Flowering
The tree continues its development by setting up increasingly vigorous branches whose structure is a succession of amphitone* and hypotone* relay axes. At this stage, the tree still has a hierarchical structure around a single large trunk. The architectural unit is established at this stage after flowering with 3 categories of axes (Table 4) and 4 orders of branching. Surveys of rural populations revealed that the first flowering occurs between the 10th and 16th year after planting. Flowering is terminal and lateral in this species and occurs in the dry season. The flower buds grow longer and invade the whole tree. Flowering occurs only on the A4 (majority) and short branches (minority). The terminal inflorescence of the axes and trunk later leads to the production of successive forks (vigorous relays), one of which is established in the extension of the trunk (vertical) until a certain point. This direction then becomes oblique and then horizontal under the effect of gravity. When the tree reaches its maximum development, the contour of the crown is rounded and irregular; the periphery of this crown is composed of sympodial structures of almost identical size and morphology. Table 4 shows the morphological and architectural description of the axis types constituting the architectural unit in Parkia biglobosa.

Older Scene
In old and aged trees, all twigs tend to droop. New sympodial twigs develop from dormant buds by piling up in the bending zones in amphitone* and epitone* positions. These structures are in fact successive partial reiterations. The total reiterations occur later, giving the impression of rescuing the tree in distress. These new reiterations regenerate the top of the tree, while the old structures fall off the tree (the tree molts). The whole tree becomes a reiterated complex; in the periphery of the crown, the sympodial structures form successive arches (piling up). At this stage, the tree can carry up to 6 orders of branching with always three types of axis categories. The branching orders are the capacity to carry an axis (carrier-to-carrier ratio), while the axis category is a set of axes groupable by common botanical entity characteristics (size, diameter, number of GUs, phytomeres, flower buds, etc.). The tree can branch up to 6 times in order, but the axis categories are repeated in the successive branching at this stage. During flowering, the reiterations also bear terminal and lateral inflorescences. Flowering takes over the whole tree at this time; the outline of the crown is rounded but irregular. Table 6 presents the analysis of variance of the morphological parameters of the modules by habitat and by locality in the sampled old trees of Parkia biglobosa. The former moduli were not influenced by the different localities surveyed in old trees (P > 0.05). However, they were influenced by habitat type (P < 0.05).
Habitat (undergrowth and full sun) altered the morphology of the growth units at the shoot tips. However, the morphology of the growth units was statistically similar in all surveyed localities (Table 6).

Relationship between the Morphology Parameters of Modules Present on Parkia biglobosa Axes Observed
The analysis of the links between the morphological parameters of the modules evaluated (Table 7) Table 8 shows the main architectural characters in Parkia biglobosa.

Discussion
Each plant follows a succession of morphological development stages throughout its life with a precise sequence ordered by different elementary entities (composition of the tree axis structure): phytomere, growth unit [5] [12] [25].
This scientific discipline has several advantages: it allows us to understand the functioning and shape of plants, to describe the biological phenomena that gave rise to them and to translate the non-linear aspect of their reaction dynamics to certain stresses [7]. In this study, the architectural analysis carried out on Parkia biglobosa along a climatic gradient made it possible to highlight its architectural characteristics. The analysis of the morphology of the modules informed us about the adaptability and the evolution of the species in relation to a changing environment. These results are an introduction to the architectural study of this species, in order to open up short, medium and long-term research perspectives.
Generally, architectural analysis on each individual is most frequently used because of its effectiveness in clearly illustrating the developmental pattern of a species. However, it is possible to analyse the three-dimensional organisation of a forest in a global way based on dendrometric data and graphical representations of plant architecture [16]. This makes it possible to develop the general architectural model of a forest [27] in order to identify the interest in forest interpretation and management or to illustrate the interest of the architectural model in the interpretation of biodiversity [16]. Indeed, the criteria used to characterise the architectural state of a plant are not always relevant and applicable to other species. It is therefore necessary to develop architectural approaches by grouping species by first studying the development sequence of key species under various stresses and behaviours in order to generalise or attribute it to a vegetation [2]. This mechanism would provide a global architectural vision of a forest for its rapid management.
Plant architecture holds many keys to understanding the ecological performance of species because resources (water, light, nutrients) are spatially variable and disturbances (e.g. frost, fire and herbivory) also impact on plants [3] [28] [29]. The main drivers of the evolution of architectural traits probably include water stress in deserts. Plant architecture strongly influences ecological performance and its role in plant evolution has recently been studied in depth by [30]

Growth Units or Modules Morphology
Parkia biglobosa is a light species and native to arid areas (savannah). However, it is shade tolerant, as the results of this study revealed that individuals observed in the nursery were taller and more vigorous than young wild individuals observed in the forest. This may be due to trauma from insect and herbivore attacks in the forest than in the nursery. Furthermore, analysis of variance showed that in young trees, the largest individuals were found in the undergrowth for all surveyed localities. This is due to the search for light; the stems elongate by means of intense apical meristematic activity induced by auxin (phytohormone) in order to reach the canopy. The majority of these trees develop fewer branches and have a tapered monopodial trunk. The majority of these trees benefit from a humid environment due to evapotranspiration of leaves from trees with higher strata. In contrast to the young wild trees, which are exposed to full sun and have less of a microclimate that is favourable to their functioning (drier soil and environment). The latter are short and very often develop reiterations that are confused with branches. Their height is smaller because the race for light is not necessary and urgent in addition to the lack of water (dry soil). The effects of light and environment on development and growth have been demonstrated in several studies [31] [32] [33] [34] [35]. The results showed that the two environments considered (undergrowth and full sun) had no influence on the morphology of the growth units in young trees. Regardless of the environment, meristem function may depend on the plant genome and therefore cannot be significantly influenced by the environment in many cases. The expressed phenotype is therefore purely related to the plant genotype. Indeed, according to [4], the genetic programme for plant growth and development, of which the architecture is the visible expression, may not vary from one environment to another in several families of species. Locality has had influences on the morphology of growth units in young trees. This is due to the difference in climate, soil type and rainfall between localities. Indeed, observations on the morphology of growth units were made along a south-north drought gradient. It is obvious that some localities have more severe (drier) conditions than others. [12] [36] and [31] have shown in their research that soil depth and fertility, environment and age of individuals can influence the architectural development of a species. [37] and [38] have indicated in their studies that climate or ecological gradient has an effect on plant morphology. Similarly, the studies of [39] [40] and [41] indicated that the origin of differences in tree morphology is due to factors such as soil type, age and genetic characteristics of individuals. They later concluded that it was possible to date a branch by counting the number of growth units or growth rings over many years with reasonable error.
However, it was still difficult to estimate the exact month of their formation in order to study climatic influences.
In old trees, the analyses of variance showed that the two habitats considered and the locations surveyed had no significant effect on the dendrometric parameters and the morphology of the growth units. This is due to the age and physiological state of these individuals. Indeed, all individuals of this age were confused due to the high intraspecific similarity. All old individuals observed in any environment had almost identical qualitative and quantitative aspects (dendrometry and morphology of growth units). At this stage the architecture can evolve, the structure of the tree is degrading and homogeneous from one individual to another. According to [4], whatever the age of the plant, young or old, the distribution of aerial and underground organs (leaves, internodes, phytomeres, growth units) is a conflict between two contradictory influences: firstly, the genetic programme for growth and development dictates architectural rules; inherent in the genome, these rules are stable and predictable in the species considered. On the other hand, adverse ecological factors (light, wind, animals, etc.), which are naturally random, often distort the architectural programme.
The Pearson matrix showed a strong positive correlation between the morphological parameters evaluated on the modules. This means that as one variable increases, the second variable also increases. For example, as module 1 gets longer, their diameter increases and the number of leaves increases. The correlation between the different dimensions of organ morphology has been demonstrated in the studies of [43] and [44]. Also, [45] and [46] obtained similar results on Tectona grandis and African coffee species respectively. However, in their studies, relationships between organ sizes of individuals from different environments were not established as was the case in our study. The allometric equations established by the linear model could be used to reduce the effort of collecting field data between the variables used and the two habitats considered if the relationships (R 2 and r) were strong (R 2 and r > 0.7). This is not the case.

Conclusions
This work firstly allowed us to understand and characterise architectural development in Parkia biglobosa and secondly to highlight the variability of morphological markers at the end of the axes in the crown of the trees. Thus, the archi- The level of organisation of the species is 5: the phytomere, the growth unit or module, the axis, the architectural unit and the whole tree (reiterated complex).
This information allows us to understand the sequential development of the overall structure of P. biglobosa. Its architectural unit is thus its smallest stem structure necessary to reach its sexual maturity stage and thus complete its life cycle by forming flowers and fruits. This study is an introduction to the architectural study of this species and the information provided will serve as a basis for further research into the architecture in relation to the sustainable use of this species.