Regeneration of Species of Natural Flora of the French Antilles Under Plantation Forests of Mahoganis (Swietenia Macrophylla and Aubrevilleana): First Data, the Case of Martinique

Background In many geographic areas inuenced by tropical and temperate climates, natural forest ecosystems have been destroyed in favour of plantations of allochthonous trees which are economically protable for different aspects of the timber industry. Some of these mature plantations degrade the soils and inhibit the regenerations of local ora species; others, due to the physical constraints which they impose, can contribute to the installation and the morphogenetic development of autochthonous taxa. The plantations of Swietenia macrophylla and Swietenia aubrevilleana (Mahoganys) in the Lesser Antilles are part of these processes. To study the regeneration methods of forest plant species native to Martinique under plantations of Mahoganys, we carried out surveys in thirteen transects (stations) inuenced by humid and subhumid bioclimates. The results showed that ane natural ora species from various stages of the plant succession colonise the plots of mature Mahoganys. This study shows that mature plantations of not very competitive introduced forest species can greatly accelerate phytocenotic succession and increase specic richness. It is therefore possible to use these introduced species (Swietenia macrophylla and Swietenia aubrevilleana) in reforestation processes of Lesser Antilles biotopes of sylvan potentiality degraded by anthropisation. Cyathea muricata, Dacryodes excelsa, Miconia trichotoma, Myrcia Fallax, Palicourea crocea, Piper dilatatum, Pouteria pallida, Talauma dodecaprtala, Tapura latifolia. The preceding factors show that the species in regeneration under the plantations of Mahoganys are little distributed. This is linked with their methods of clustering or "socialisation" ranging from dispersed to gregarious and the highly differentiated quantitative signicance of their populations. The reasons for this appear to originate from the strategies of dissemination of seeds or diaspores. For the dispersal of their seeds and in a proportional manner, the regenerating species favour zoochory (notably ornithochory) and barochory. However, in the case of trees which are among the most numerous physiognomic types, barochory conditions higher densities of gregarious populations when the following two categories of autochthonous seeders are present: relictual trees with low growth found within or at the edge of Mahoganys plantations which belonged to ancient natural forest groupings, and adult trees with average or fast growth which underwent their morphogenetic development within the Mahoganys plantations. excelsa, Pouteria multiora, Pouteria pallida, Sloanea dendata, Sloanea dussii, Sloanea massoni, Talauma dodecapetala and Tapura latifolia are predominantly matrix species of mature pre-climactic and climactic forest groupings of humid and humid subhumid bioclimates. Pimenta racemosa is a dominant taxon [high distribution (Id) and high overall basal area] of the advanced phases of the phytocenotic succession of pre-climactic and climactic forest formations inuenced by the dry subhumid bioclimate. The other taxa listed are characteristic of various types of sylvatic gaps and severely degraded forests which are called regressive (young and secondary forests, see Annex 2). Rather pertinently, this study shows that the Mahoganys plantations can be used as ecological engineering tools to activate plant dynamics, particularly in highly diminished or unstructured environments with humid, humid subhumid and dry subhumid bioclimates. Anthropisation, as well as the frequent cyclonic hazards, leads to a regression from the forest stage to the herbaceous, bush or mixed stage of the advanced, pre-climactic and climactic plant ecosystems. In this respect, the requirements of forestry, in particular to facilitate the cutting and extraction of logs (trunks), should not be imposed as regards the distances between the specimens of the old Mahoganys plantations, which are too great. This results in a phase difference between the macroclimate and the intra-plantation climate which can be medium to low depending on the level of overall morphogenetic development of the individuals. The dense plantation of Mahoganys would be the most logical solution because at the end of the growth of individuals, shorter exclusion distances between them lead to an interior environment that is climatically much more dephased climatically. In general, this study shows that mature plantations of not very competitive introduced forest species can greatly accelerate phytocenotic succession and increase specic richness. (Richards et al. 2010; Bremer and Farley 2010; de Jong 2010; Paquette and Messier 2010; Verheyen et al. 2016; Aide et al. 2013).


Abstract
Background In many geographic areas in uenced by tropical and temperate climates, natural forest ecosystems have been destroyed in favour of plantations of allochthonous trees which are economically pro table for different aspects of the timber industry. Some of these mature plantations degrade the soils and inhibit the regenerations of local ora species; others, due to the physical constraints which they impose, can contribute to the installation and the morphogenetic development of autochthonous taxa. The plantations of Swietenia macrophylla and Swietenia aubrevilleana (Mahoganys) in the Lesser Antilles are part of these processes.

Methods
To study the regeneration methods of forest plant species native to Martinique under plantations of Mahoganys, we carried out surveys in thirteen transects (stations) in uenced by humid and subhumid bioclimates.

Results
The results showed that a ne natural ora species from various stages of the plant succession colonise the plots of mature Mahoganys.

Conclusions
This study shows that mature plantations of not very competitive introduced forest species can greatly accelerate phytocenotic succession and increase speci c richness. It is therefore possible to use these introduced species (Swietenia macrophylla and Swietenia aubrevilleana) in reforestation processes of Lesser Antilles biotopes of sylvan potentiality degraded by anthropisation.

Background
The domestication of species for survival and societal development is common to all civilisations (Hawkes, Maxted and Ford-Lloyd 2000;Gremillion 1997;Terrell at al. 2003). These species have enabled different peoples to ensure their longevity in all living environments (Crawford 2006; Bar-Yosef 2017; Simons and Leakey 2004). Over ten thousand years ago, the neolithic revolution linked to the birth of agriculture resulting from the gradual or accidental domestication of plants and animals was a turning point in the evolution of humanity (Smith 1997 Stehlé 1957;Howard 1995;Weaver, Valenta and O cer 1985). In Martinique, the National Forestry O ce (ONF in the French acronym) oversees the entire Mahoganis production chain and for more than a century, Mahoganis plantations have been carried out throughout Martinique. Currently, certain plots of Mahoganis over 40 years old are not exploited and their crowns bring about particular internal eco-climatic conditions. In this study, the research protocol is based on trasnsects located in different plots of Mahoganis in order to inventory the populations of autochthonous species (Fig. 2). It is ultimately a question of showing the role of the structure of these mature Mahoganis plantation forests in the installation and development of species of native ligneous ora, especially those from the advanced stages of plant succession. What are the mesological factors that condition the regeneration of extra-and intra-forest autochthonous ligneous species? What are the distribution methods of the species populations? What are the causes of the qualitative stational differentiation of the species?

Materials
General biophysical characteristics Martinique, like the other Lesser Antilles, stems from an intra-oceanic subduction which gradually constructed its contrasting geomorphology (Garmon, Allen, and Groom 2017; Costa, Audin and Benavente 2009). The different topographical facies that result from it modify the structure of the main climatic factors and contribute to the considerable diversity of the biotopes (Triantis et al. 2003). Orographic rain is the key factor of the plant colonisation (DeWalt, Ickes and James 2016; Howard 1962). This leads to, from the lowlands to the mountain peaks, bioclimates which differ from the altitudinal point of view between the windward slope (eastern slope) and the leeward slope (western slope). This bioclimatic staging is a staging of phytocenotic potentiality. Indeed, from the littoral to the altitudes where sylva can install themselves, we nd in particular: the lower tropical seasonal evergreen forest in its most xeric facies, the typical tropical seasonal evergreen forest, the sub-montane tropical ombrophilous forest and the montane tropical ombrophilous forest. Ecotones allow for the transition between these potential sylvan types. Generally speaking, and regardless of the plant stages, which are therefore bioclimatic, the plant cover is a patchwork of phytocenoses of different structure, speci c composition and ages. Originally, species of the genus Swietenia from Martinique were planted in the bioclimates mentioned above for the production of precious wood.
Main botanical and phytogeographical characteristics of the species of the genus Swietenia. From a phyllogenetic point of view and in ascending manner, the genus Swietenia (Jacq.) belongs to the Family of Meliaceae, the Order of Sapindales, the Class of Magnoliopsida (Dycotyledon), the Division of Angiosperms (Magnoliophyta) and the Kingdom Plantae. The Genus Swietenia comprises ve species, all native to tropical America (Krisnawati, Kallio and Kanninen 2011;Snook 1996, York 2012; Fig. 3). However, four of them have been introduced into numerous geographic regions of the intertropical zone [Castañeda-Posadas and Cevallos-Ferriz 2007 (Fig. 3)]. Swietenia aubrevilleana is the result of a hybridisation between macrophylla and mahagoni and is present mainly in certain islands of the archipelago of the Antilles (Brown, Jennings and Clements 2003;Gleason and Panshin 1936;Mabberley 1982). Swietenia mahagoni, macrophylla and aubrevilleana are the only species that are present in Martinique (Fig. 4). From the physiognomic point of view, they have a strong resemblance, but aspects pertaining to the foliar unit and the owers easily differentiate them (Fig. 4). In adulthood these trees can peak at 40 metres in height with sections at 1.33 metres from the ground ranging from 1.5 to 2.5 metres and ensuring their morphogenetic development on soils conditioned by humid, humid subhumid and dry subhumid bioclimates (Fig. 4).
Study stations, mesological factors, potential forest types and stages of evolution.
The oristic survey stations were chosen from within mature Mahoganis plantation forests and are located along an altitudinal gradient from 258 m to 400 m. They are under the in uence of hyper-humid, humid and subhumid bioclimates (Table 1a) corresponding to a tiering of mesological factors which condition particularly tropical montane, tropical sub-montane and tropical seasonal evergreen phytocenoses. Associated with the different types of vegetation are soils such as andosols, fersiallitic soils and vertisols (Tables 1a and b). The average annual temperatures measured using meteorological stations close to the oristic survey units (Fig. 2) vary between 23 °C and 25 °C with minimums and maximums of between 20 °C and 23 °C and between 27 °C and 29 °C (Tables 1a and b). The annual rainfall, which is the factor which characterises the bioclimates by in uencing the types of plant ecosystems, varies between 2613.1 mm and 3869.6 mm with signi cant seasonal differences (Tables 1a and b). The number of days with rain during the wet and dry period indicates that the study stations are in rainy areas (Tables 1a and b). The latter are characterised by annual insolation levels and daily global radiation levels which are among the lowest in Martinique (Tables 1a and b). These aforementioned features are consubstantial with a high annual maximum humidity (100%) and with an annual evapotranspiration of between 1550 mm and 1600 mm.

Methods
The methodology used for this study is based on macroecology. The objective is to decipher the structural and functional dimensions of the vegetation. Using transects subdivided into quadrats and according to the minimum survey area (between 250 to 1050 m²; Table 1b) we generated data which are both average ecological and oristic descriptors ( Fig. 5): species, numbers of individuals of populations of plant species [from regenerations to mature specimens (bio-demographic aspects)], diametric classes, classes of total heights and rst branchings.
These features allowed us: -to assess the distribution of the sections of the individuals, the architecture of the formations, the characteristics of the canopies.
-to evaluate the phytomasses or biovolumes using the basal area (St) which corresponds to the sum of the areas of the circles that constitute the sections measured at 1.33 metres from the ground in accordance with international standards.
-to nd out the distribution of the species between the quadrats of the transects and between the stations using the Index of distribution which corresponds to the following formula: Id = fr × d (fr being the relative frequency and d (nb / Sr) the density corresponding to the number of individuals of the species (nb) divided by the survey area (Sr , Table 1b).
-to discover the relative dominance of ligneous plant species in relation to each other via the Index of Dominance (ID). ID = Id × St (basal area).
-to conduct a Correspondence Factor Analysis (CFA) and an Ascending Hierarchical Classi cation (AHC) from the XLSTAT software (new version), which allowed us to compare the stations with one another with regard to the population structures of the different species.
-to conduct an analysis of the species by means of clustering taking as a reference the fr (relative frequency) and Density [two features of the Id (Index of distribution)]. This clustering was done using the k-means algorithm (Hartigan and Wong 1979). Variables were normalised.

Results
Taxonomic diversity of natural ora in regenerationunder Mahoganys plantation Speci c diversity is also signi cant, but with respect to the study plots, a small number of species have a Relative Frequency (FR, Fig. 8a   Under the plantation forests of the study stations, the set of the epigeic ligneous individuals with bush and tree potentiality presenting various morphogenetic phases mostly have, at 1.33 cm from the ground, section diameters of 2.5 cm (Fig. 9a). From the quantitative point of view, the reduction factor is 3.74 among the populations of autochthonous species of the 2.5 cm and 5 cm classes, while the reduction factors are 13.9 and 21.8 respectively among those of the 2.5 cm, 10 cm and 15 cm classes. As shown in Fig. 9a, for classes with a diameter greater than 15 cm the number of stems is between 39 and 663 times lower than that of the 2.5 cm class. The same is true for heights and rst branchings (Fig. 9b)
The distribution of species populations with respect to height class seems to follow the same trends as that observed for diameter class. Indeed, the majority of individuals have a height of between 1 m and 8 m and correspond to a number of taxa, particularly those whose number of specimens is greater than 50 (Fig. 11). Chimarrhis cymosa, Cyathea arborea, Cyathea muricata, Dacryodes excelsa, Guatteria caribaea, Heliconia caribaea, Myrcia fallax, Myrcia splendens, Pouteria pallida are the species which differ from those of the majority group considered with respect to the 2.5 cm diameter class with at least 50 individuals (Fig. 11). Logically, the majority of the rst branchings are between 1 m and 8 m (Fig. 12). However, quantitatively low individuals of a few species have their rst branchings in greater height classes (Fig. 12). There seems to be a fairly signi cant relationship between the life history of the species, in particular their morphogenetic development, and their populational signi cance relative to the classes of diameters, heights and rst branchings.

Distribution and ecological dominance of species for all stations
The Index of distribution (Id) of the species vis-à-vis the thirteen oristic survey stations results from the product of their relative frequency (FR) and the density (d) of their individuals (see methods section). This Id is rather signi cant evidence of the mode of populational distribution of the various regenerating species with respect to the inventory areas and indicates types of sociability ranging from dispersed to gregarious. Figure 13 shows an interspecies variation of the three aforementioned indicators. There seems to be a certain alignment of the curves. The Relative Frequency (FR) curve indicates that the majority of species are on average present in 30% of the stations, i.e. around four. The taxa whose relative frequency is greater than 60% are the following: Aniba bracteata, Chimarrhis cymasa, Cestrum laurifolium, Cyathea arborea, Cyathea muricata Dacryodes excelsa, Miconia trichotoma, Myrcia fallax, Palicourea crocea, Piper dilatatum, Pouteria pallida, Simarouba amara, Talauma dodecapetala, Tapura latifolia. In general, apart from Tapura latifolia whose density of individuals is relatively higher, the density of regenerating populations is low or even very low for certain species (Fig. 13). This leads to Indices of distribution (Id) which are also low to very low. At the level of the thirteen inventory plots, the most distributed species with an Id greater than or equal to 0.015 are, in order of signi cance: Tapura latifolia, Odontonema nitidum, Myrcia fallax, Myrcia spledens, Palicourea crocea, Psychotria muscosa, Piper dilatatum, Pouteria multi ora, Trichilia septentrionalis (Fig. 13). Figure 14 shows, for all the stations, four clusters based on the presence of the species (FR) and the size of their population (density).
Despite the low distribution of species characterised by the Id, it was necessary to determine their ecological signi cance using the Index of Dominance (ID). The latter results from the product of the Basal Area which is the sum of the cross-sectional areas of the stems of trees and bushes of each species at 1.33 m and of the Id. Tapura latifolia, Chimarrhis cymosa and Dacryodes excelsa are, in order of signi cance, the three dominant species with respect to the set of survey stations (Fig. 15).

Multifactorial analyses
The Correspondence Factor Analysis (CFA) based on a contingency table composed of 136 rows (species) and 13 columns (inventory stations) shows that there is a link between the rows and the columns since the p-value (0.0001) calculated is below the alpha signi cance level (0.05). Axes F1 and F2 (Inertia: 38.70%) of Fig. 16a distinguish between stations both from a bioclimatic point of view and from a taxonomic point of view: stations with humid bioclimates [SM1, SM2, GM1, GM2, La1, La2, Ab1, Ab2, Ab3, TM, MT (Table 1)]; stations with subhumid bioclimates [FL1, FL2 (Table 1)]. The different groups of species are a ne to the following ecosystemic potentialities (Fig. 16b): tropical sub-montane ombrophilous forest (Group A), tropical ombro-evergreen forest (Group D), tropical evergreen forest [Group B (Sloping area) and Group C (Riparian terrace)]. The distances between the species segregated by the rst two factorial axes correspond to various degrees of similarity of their biodemographic structures. Within groups A, B, C and D, the Relative Frequencies (FR) and the densities (d) of the species are appreciably similar while between these groups the speci c differences (of the taxa) with regard to these indicators (FR and d) are pronounced. Pp (Picramnia pentandra) is a species common to stations Fl2 and La-2 / Cly (Conostegia calyptatra) is a species in regeneration common to stations G-M2, Ab3 and Fl2 The FL1, FL2 and Mt stations differ from the others from a oristic point of view, and so we have removed them from the contingency table. A new Correspondence Factor Analysis concerning only the stations in uenced by the humid bioclimate (Group A, Fig. 16a) was carried out based on the new contingency table of 89 rows (Species) and 10 columns (Stations). The different positions of these humid stations on the plane de ned by the rst two factorial axes of Fig. 17a are related to their corteges of species in regeneration. It seems that each station is characterised by a phytocenosis of which certain taxa are highly speci c. In reality, the degrees of taxonomic difference correspond to more or less signi cant distances between the stations presented on the plane formed by the factorial axes F1 and F2 (Fig. 17a). As for the variations in distance between taxa (Fig. 17b), they result from their differences in populational structures. These are related to the minimum survey areas ( Table 1). The combinations of species (I, II, III, IV, V, Fig. 17b) which distinguish the study stations are just particular phytocenotic expressions of the tropical ombro-evergreen and submontane ombrophilous forest vegetation. Two Ascending Hierarchical Classi cations (AHC) based on general similarity were carried out based on the set of the stations (13) and their species (136) highlighting many distinguishing features both between stations and between species. First, the set of the stations have a similarity of 0.434. Two groups of stations subject to the spatiotemporal dynamics of humid (group A) and humid subhumid (group B) bioclimates stand out (Fig. 18). They are each subdivided into two subgroups (A1, A2) and (B1, B2) whose similarity is 0.58 (A1, A2) and 0.53 (B1, B2). The stations in subgroup A1 present various degrees of similarity ranging on average from 0.67 to 0.82. This process is identical for stations in subgroup A2, except that the degrees of similarity are lower (Fig. 18). Overall the degrees of similarity between the stations of group B are lower than those of group A. The ecosystemic differentiations are clearly explained in Fig. 18 and the stations with the most similarity are Sainte Marie 1 and 2. Second, the Ascending Hierarchical Classi cation carried out based on the populational structures of the species segregate three groups (I, II, III) which are completely separated and therefore have no similarity (similarity index equal to zero). Species in groups I and II have just about average to very low similarity in population structure between 0.075 and 0.46 (Fig. 19). Group III, made up of two subgroups (IIIa and IIIb), includes the majority of the species whose degrees of populational similarity vary from 0.075 (IIIa: a single species) to 1 (Fig. 19). Figure 19 shows a series of subdivisions, from the lowest value  Fig. 7] are typical of the sub-montane ombrophilous and seasonal ombro-evergreen oristic formations. The same is true for the species with respect to the bioclimates and types of vegetation mentioned above, except that the latter are associated with different stages of phytocenotic succession (Fig. 22). Table 2 shows the most frequent species in regeneration [FR FR ≥ 50%] (See Annexes 1 and 2).  Ecological signi cance of the species (all stations) The index of distribution used in this study is a global index since the relative frequency (FR) and density (d) are global indicators relative to the set of stations. This index of distribution (Id) is highly variable and results from the covariant viability of the Relative Frequency (FR) of the species and the density (d) of their individuals (Fig. 13). Indeed, the number of individuals varies both between the quadrats of the transects making up the survey stations and between the latter (the transects). Generally speaking, species in regeneration have a cumulative density of individuals (for the thirteen stations) which is low or very low, although this seems to be strongly correlated with the relative frequency (Fig. 13). Within each inventory transect (station), the absolute frequency or number of species occurrences with respect to the quadrats varies as well as their populations. This process is similar between stations. The most widely distributed species correspond to those whose FR ≥ 69 and d ≥ 0.005 (Fig. 13): Aniba bracteata, Cestrum laurifolium, Chimarrhis cymosa, Cythea arborea, Cyathea muricata, Dacryodes excelsa, Miconia trichotoma, Myrcia Fallax, Palicourea crocea, Piper dilatatum, Pouteria pallida, Talauma dodecaprtala, Tapura latifolia. The preceding factors show that the species in regeneration under the plantations of Mahoganys are little distributed. This is linked with their methods of clustering or "socialisation" ranging from dispersed to gregarious and the highly differentiated quantitative signi cance of their populations. The reasons for this appear to originate from the strategies of dissemination of seeds or diaspores. For the dispersal of their seeds and in a proportional manner, the regenerating species favour zoochory (notably ornithochory) and barochory. However, in the case of trees which are among the most numerous physiognomic types, barochory conditions higher densities of gregarious populations when the following two categories of autochthonous seeders are present: relictual trees with low growth found within or at the edge of Mahoganys plantations which belonged to ancient natural forest groupings, and adult trees with average or fast growth which underwent their morphogenetic development within the Mahoganys plantations.
As the results of the Correspondence Factorial Analyses (CFA) and of the Ascending Hierarchical Classi cations (AHC) show, the mesological factors and the dissemination vectors as well as the phenological methods of the seeders adjoining the stations are determining factors which condition the structure of populations of regenerating species. These are a ne of humid, humid subhumid and dry subhumid bioclimates and consist of submontane ombrophilous, seasonal ombro-evergreen and seasonal evergreen phytocenoses respectively. When the individuals of the taxa are small in number and distributed in a large number of stations, one may think that ornithochore dissemination is the most effective. However, when there are a large number of individuals and they are distributed in a low number of stations, the predominance of barochory seems to be the most plausible explanation. On the other hand, there may be situations where the species have low relative frequencies with sparse populations. In the latter case, it is very likely that these species are largely barochoric and have phenological peculiarities resulting in infructescences with little fruit. The phase difference between the external environment (macroclimate) and the internal environment of these "Mahogany plantation" biosystems due to a very simpli ed strati cation would be in covariance with multiple sites of installation and expansion of fairly speci c factorial characteristics. Indeed, in these forests which one could call " elds of planted trees" the very simpli ed strati cation is composed of a dominant upper stratum of adult Mahoganys, a middle stratum formed of scattered individuals (therefore sparse) of expanding morphogenetic Mahoganys and mature autochthonous trees. This state of affairs, associated with the de ciency of the disseminating vector fauna, could constitute a clue that would explain the low population densities of the vast majority of the species.

Conclusion
Despite their non-contiguous crowns forming a sparse canopy, the plantations of Mahoganys, due to the geometrical constraints which they impose, enable the colonisation of species of natural ora. The oristic differences observed between the inventory stations largely depend on the seed dispersal processes and the structure of the mesological factors. Among these are the stational bioclimates, the declivity of the grounds and the types of edaphic systems. There are also other parameters such as the quantitative signi cance of the seeders adjoining the planted plots and the frequency of ooding and exposure of low-lying riparian terraces. Regenerating species present a broad taxonomic spectrum and belong to stages of the plant dynamic ranging from pioneer stages to climactic stages through to post-pioneer stages (Fig. 22). The fact that the species of advanced phases (pre-climactic and climactic) regenerate under mature plantations of Mahoganys constitutes a shunt to the dynamics of the vegetation. Indeed, species such as Aniba bracteata, Guatteria caribaea, Chimarrhis cymosa, Dacryodes excelsa, Pouteria multi ora, Pouteria pallida, Sloanea dendata, Sloanea dussii, Sloanea massoni, Talauma dodecapetala and Tapura latifolia are predominantly matrix species of mature pre-climactic and climactic forest groupings of humid and humid subhumid bioclimates. Pimenta racemosa is a dominant taxon [high distribution (Id) and high overall basal area] of the advanced phases of the phytocenotic succession of pre-climactic and climactic forest formations in uenced by the dry subhumid bioclimate. The other taxa listed are characteristic of various types of sylvatic gaps and severely degraded forests which are called regressive (young and secondary forests, see Annex 2). Rather pertinently, this study shows that the Mahoganys plantations can be used as ecological engineering tools to activate plant dynamics, particularly in highly diminished or unstructured environments with humid, humid subhumid and dry subhumid bioclimates. Anthropisation, as well as the frequent cyclonic hazards, leads to a regression from the forest stage to the herbaceous, bush or mixed stage of the advanced, pre-climactic and climactic plant ecosystems. In this respect, the requirements of forestry, in particular to facilitate the cutting and extraction of logs (trunks), should not be imposed as regards the distances between the specimens of the old Mahoganys plantations, which are too great. This results in a phase difference between the macroclimate and the intra-plantation climate which can be medium to low depending on the level of overall morphogenetic development of the individuals. The dense plantation of Mahoganys would be the most logical solution because at the end of the growth of individuals, shorter exclusion distances between them lead to an interior environment that is climatically much more dephased climatically. In general, this study shows that mature plantations of not very competitive introduced forest species can greatly accelerate phytocenotic succession and increase speci c richness.

Declarations
Ethics approval and consent to participate Not applicable. Figure 1 The Lesser Antilles in the Caribbean.

Figure 2
Survey areas and GPS coordinates (X and Y).

Figure 3
Speci c taxonomy and distribution.

Figure 4
Some eco-climatic, oristic and evaluation features. Unit of oristic surveys and these subdivisions.         Socio-biological differentiations of the species (See Annex 1).

Figure 15
Speci c variation of the basal area, the index of distribution and the index of dominance (trees and bushes).

Figure 16
A: Bioclimatic a nities of the stations (see Table 1). B: Ecosystemic a nity of the species (Box1). Biodemographic similarity of the stations.

Figure 19
Populational similarity of the species (see Box 1).

Figure 20
Relictual adult trees from old forest matrices or windthrow. Figure 21