Effects of best land management practices on the structure, composition and diversity of woody vegetation in agroforestry parklands: case of Dan Saga and Tabofatt villages in Niger republic

The large-scale use of best land management practices has been instrumental in the accession of recent regreening in certain Sahelian areas. This study therefore aimed at highlighting the effect of the wide use of these practices on the structural and ecological parameters of the woody flora in agroforestry parklands. To this end, an inventory of ligneous vegetation was carried out on radial transects in an East–West and North–South direction in the villages clusters of Dan Saga and Tabofatt in Niger Republic. A variance comparison test was applied on the values from plots with practice compare to those from plots without practice (controls). In Dan Saga site, the results obtained show the significant effect of the practice of assisted natural tree regeneration on the density of trees (97 against 58 plants/ha for the control plots), on the basal area (3.04 against 1.98 m2/ha for the control plots), on regeneration (61.96% against 40.91% for the controls) and on floristic diversity (2.93 against 2.41 bits for the control plots) with a clear dominance of Combretaceae. On the other hand, this effect is relatively less significant in the case of Tabofatt site’s where the practice of planting trees coupled with water and runoff collection structures induced positive effect particularly on the basal area (2.29 against 1.90 m2/ha for plots controls), on the structure of the vegetation stood by the abundance of plants with average diameter (class from 5 to 10 cm) as well as the species diversity (2.54 against 2.18 bits for the controls) with a dominance of species of the Mimosaceae family. These results can be used as decision support tools to guide actions to scale up regreening in the Sahelian context.


Introduction
The Sahel was shaken by major ecological crises during the 1970s and 1980s which deeply affected the agrarian landscape. In some cases, in response to these crises, Sahelian populations have invested enormous efforts that have led to a return of vegetation [1][2][3]. This landscape revegetation process has resulted in the establishment of agroforestry parklands. According to [4] Agroforestry parkland is a land-use system in which woody perennials are deliberately conserved in association with crops and/or animals in a dispersed spatial arrangement and where there are both ecological and economic interactions between woody and the other components of the system. This process commonly called "Sahelian re-greening" has been well documented by several studies using remote sensing and GIS techniques. Furthermore, it has 1 3 been observed that where human concentration is high, land degradation is low compared to areas of high land availability [5][6][7]. The large-scale adoption of best land management practices has played a decisive role in the trend towards landscapes re-greening in the heavily anthropized areas [5,8,9]. The best land use practices includes land management practices used to improve land productivity and ecological restoration. They are mostly used to restore degraded lands in the sahelian region. These practices are affordable simple for peasant and environmental friendly such as half-moon, stone line, tree regeneration farming, assisted tree regeneration, and so one [10].
Regarding to the local specificities, many land restoration stakeholders has designed technical reference which played a guide in order to better adapt their approaches. In this perspective, a few success stories had emerged in Niger which induced a tremendous process of the vegetation recovery on the landscapes. On this line many authors highlight the roles of the practices of Assisted Natural Trees Regeneration (ANTR) on sandy soil and the construction of soil erosion and run-off structures coupled with planting trees on encrusted glacis [5,11,12]. For instance the ANTR, one of the best land use management is an indigenous practice widely adopted by rural population in certain area of the Sahel region. In this practice farmers spare new growth during the farming activities. These young tree shoots are protected and maintained by trees breeding actions until their maturity during which they are able to produce enough wood, fodder and other non-timber forest products. This allows for the continual renewal of the woody cover which results in the strengthening of community resilience and the biodiversity of the landscape. This practice has contributed in the re-greening of more than 5 million ha of degraded land in south central Niger region [6].
Several insights on parklands woody flora are fairly well documented. The previous studies were restrictive on certain specific ecosystem or some species within the landscape. For instances [13] provided the spatial repartition and demographic structure of Sclerocarya birrea (A. Rich;) Hochst. in the Niger sahelian sector (Dan Saga). [14] described the ecological and structural pattern of Maggia vegetation (Tabofatt). However there is still a need of enough understanding on how did land use practices influence the structure, composition and diversity of woody vegetation. That is critical in the acknowledgment of the landscape dynamic. Thus, the purpose of this study is to highlight the effects of best land management practices on the structure composition and diversity of the woody flora of two agroforestry parklands where the use of practices has resulted in vegetation cover enhancement. This study aims at pointing out some specific guide lines which could play a key role in the scaling up of re-greening process in the context of Great Green Wall Initiative of the Sahel.

Study area
The criteria used in the choice of the study area were based on the high population density (over 100 inhabitants per km 2 ) and the high level of appropriation of best land management practices. In addition, to discriminate the types of practices used according to local soil and climate characteristics, the study was performed in the way to bring out the evidence of ANTR (in one hand) and tree planting coupled with the erosion and runoff structures (on the other hand) in the expression of re-greening process. It is on this basis that the village clusters of Dan Saga and of Tabofatt were chosen to serve as a study area.
The Dan Saga cluster is located in the Aguié department (Maradi region, south-central Niger). It covers 7 villages namely: Dan Saga, Gardjagaou, Nakikarfi, Guidan Gaida, Dan Dawa, Dogaraoua, Guidan Bakoye Sabon Gari, circumscribed between geographical coordinates 13°30' and 13°50' North latitude and 7°30' and 7°60' West longitude. The climate is dry tropical [15]. According to Saidou et al., the last trend of the rainfall regime stated from 1994. The average annual rainfall recorded on the time series 1994-2018 is 485.19 ± 71.71 mm with a variation coefficient of 14.76%.The relief is in the form of a vast relatively undulating dune plateau with leached or slightly leached tropical ferruginous soils [16]. These are deep soils (1.5 m on average) with a particulate structure and texture dominated by coarse and medium sands at depth and more or less fine in the superficial horizons. This predisposes this type of soil to infiltration and therefore a low water retention capacity. Due to these pedological and climatic features local farmers have adopted in their farming system the practice of Assisted Natural Tree Regeneration (ANTR) [17]. The ANTR is an endogenous practice that consists of identifying, selecting, maintaining and protecting young shoots from woody species in the fields. Its potential is determined by the stock of seeds disseminated in the soil, and the release of stumps [8,10,18].

3
Tacharou, Tikirana Taibaroug and Ouroub. It is located between 14°40' and 14°50' North latitude and 5°50' and 6°00' West longitude. The north-Sahelian position places the cluster in a sub-arid climate with low annual rainfall. As an illustration, Saidou et al. found an average annual rainfall of 359.14 ± 62.21 mm with a variation coefficient of 17.32% over the time series from 1988 to 2018. The showers are often of high intensity: more than 50 to 60 mm per hour [19]. The geomorphology is in the form of a sandstone plateaus cut by deep valleys (Maggia). Typical brown soils and sub-arid red-brown soils occupy most of these cultivated glacis [20]. On the top of the sandstone plateaus, the soils are thin with a clayey texture favorable to good water retention. In the valleys, the depth of the soil is very variable depending on the topography. For example, clayey lowland soils are favorable for flood recession crops, while glacis soils can only be farmed during the rainy season. In this context, the populations have largely adopted the practices of half-moon, zaï and stone lines coupled with the planting of trees in the fields. These practices aim to collect erosion and runoff that will be returned to the young plant on site.

Material
The characterization of the ligneous vegetation of the two sites was made using a batch of equipment intended for the realization of the ligneous flora inventory. Each instrument was used in a specific way (Table 1).

Data collection
Field data collection was done on the basis of systematic sampling on East-West and North-South radial transects [21]. The survey lines intersect at the most central village. The inventory units are plots of 50 m*50 m (i.e. 2500 m 2 ), which correspond to the minimum area generally used during inventories in agroforestry parklands in West Africa in the case of woody species [22]. These plots were arranged along survey lines with an equidistance of 200 m (Fig. 1). Regeneration was inventoried in plots of 5 m*5 m (25 m 2 ) inside the survey plots [23]. In all directions, the first inventory plot is placed 500 m from the village to avoid the influence of dwellings.
Within each plot, the data collection focused on: (i) dendrometric measurements of all woody plants including shoots, (ii) parameters related to station conditions (geographic coordinates and altitude of the central point, texture of the soil) and (iii) type of best land management practice found within the plot. In each of the two cases, the survey was conducted taking care to clearly discriminate between the modality with practices (Picture 1) vs control plot. In the case of ANTR, the modality with practice is noticed when two or three rejections of stumps are spared following farmer care on the tree. With regard to erosion and runoff collection structures, the survey focus on plots with planted tree half-moons, planted tree stone lines, represented the case of plot with practice. Data collection took place during the month of August 2018. In Dan Saga cluster, 61 plots were recorded, including 39 plots with ANTR and 22 control plots. For the Tabofatt cluster, 25 plots were recorded, divided into 13 plots with practice and 12 control plots.

Descriptions of structural and ecological parameters of woody plants
The description of the structural characteristics focused on the basal area (1), the tree density (2), the regeneration rate (percentage of the number of individuals with a diameter less than 5 cm over the number of individuals with a diameter greater than 5 cm) and the stand renewal rate (percentage of the number of plants with a diameter less than 5 cm relative over the total number of plants) [24].
G is the basal area (m 2 .ha −1 ), D i is the diameter at breast height of tree trunk (m) i: specie and N total plots number.
Here n represented the total number of tree recorded in a given plot. The structure of woody vegetation was visualized on the basis of the distribution of ligneous plants frequencies in function class diameter. This distribution was fitted to the theoretical Weibull density function with 3 parameters a, b and c (3).
In this formula, x denotes the diameter of the trees, a: the position parameter, b is the scale or size parameter and c expresses the shape of the structure [25]. In this study, we retained the shape parameter c to deduce the evolution of the woody population.

Fig. 1 Woody vegetation inventory frame. VC Village central, P Plot
The description of the floristic procession was made according to the calculation of the importance value index of each species (IVI) (4). The identification of the species was made with reference to Peyre de Fabregues (1979), [26] and the national herbarium of Niger, available at the Plant Biology Mounkaila Garba Laboratory of the Faculty of Sciences and Techniques, Abdou Moumouni University of Niamey.
The relative dominance (expressed in %) of a woody species is the quotient of the total basal area of the given specie in the stand by the total basal area of all the recorded species. The relative density of a species is the quotient of the total number of plants species over the total number of plants in the area. The relative frequency of the species express in percentage refers to the frequency of the species (number of plots where the species is present over the total number of plots) by total frequencies of all species.
The description of the intra diversity parameters focused on: the total richness defined by the number of listed species, the Shannon-Weaver diversity index α (5) and the Piélou even-distribution index (6).
where ni denotes the number of tree species i in the plot and n the total number of tree in the plot. pi is the relative abundance of species i in a given plot.
S refers to the total number of species recorded in a given plot.
The inter diversity is highlighted by the Sorensen Index (7).
Is is the Sorensen Index, c is the number of species recorded in both cases (plots with practice and control plots), a is the number of species recorded in plots with practice and b in the one in control plots. It expresses the level of similarity between plots with practices and control plots.

Statistical analysis
The statistical analyzes aimed at comparing the means value between the situation with and without practices. To this end, at the level of each plot, the values of the structural and ecological parameters were determined using the Excel spreadsheet. The two-sample variance comparison module of the XLSTAT software was used to generate and compare the statistical means of the series of values with practices and without practices within the two agroforestry parklands. The level of significance of the results of this analysis was determined using Fisher's test at the 5% level.
Pearson's Chi-square adjustment test was used to figure out the level of adjustment between the observed and theoretical Weibull values. This aimed to assess the accuracy of the fit between the two data series.

Effects of best land management practices on the structural characteristics of woody stands
The values of the structural and evolution parameters (Density of Trees, Basal Area, Regeneration Rate and Tree Renewal Rate) of recorded value are listed in Table 2.
At Dan Saga, the results show that the density of trees in plots with ATNR is significantly higher (97 individuals/ ha) compared to control case (38 individuals/ha). At Tabofatt, tree density of control plots is slightly higher than the one in plots with practice (93 against 86 individuals/ha). In both cases, the values of the basal diameter of the plots with practices are clearly higher than the controls. In terms of evolution the highest value of the regeneration rate is (4) Importance Value Index = Relative Dominance + Relative Density + Relative Frequency.
recorded in plots with practice (for instance 61% in plots with practice) compared to control plots 40% in the case of Dan Saga). However, it is easy to notice that the standard deviation of all these values is strong. In the case of Tabofatt, the regeneration rate of control plots is quite higher than those of plots with practice. Both in Dan Saga and Tabofatt, the values of Tree Renewal Regeneration are not significantly different between plots with practice and control plots.
In over all cases, the results of the Chi-2 adjustment test of the tree density distribution of observed and Weibull theoretical values in function with the class diameter are significantly well adjusted (value of p-value ≥ 0.0001). This allows an easy interpretation of the woody vegetation structures. The diametrical structure of the woody population of the Dan Saga cluster presents the same appearance in the case of plots with practices as well as the control case ( Fig. 2a and b). In both cases, the theoretical Weibull distribution presents an inverted J trend with a shape parameter value substantially close to 1 (c varying from 1.09 for the case of the plots with practice and 1.12 for the case of the On each parameter, the letter a assigned to the value of modality with practice and without practice means that the difference is not significant. If the letter a was assigned to the modality with practice and b to the modality without practice, that means the difference is statically significant at the 5% level  control plots). However, it is noticeable that the Weibull function of density trend is strictly monotonous in the case of the control plots.
In the case of the Tabofatt cluster, the diametric class structure of the plots with practice displays a truncated bell shape on the left. The woody vegetation structure of the control plots globally presents an inverted J trend ( Fig. 3a and  b). The values of the shape parameter are clearly greater than 1 with c = 1.98 for the plots with practice and c = 1.47 for the control plots.

Effects of best land management practices on the ecological parameters of woody vegetation
The specific composition of the woody vegetation of plots with or without practice is presented in Tables 3 and 4  The average values of the diversity parameters per plot are provided in Table 3. They serve as references to compare the effect of the land management practices on the floristic diversity of the woody vegetation.
The results in Table 3 show an average of 10 species per plot with practice and 7 species per control plots for the case of the Dan Saga site. They also show a much lower specific richness on the Tabofatt site where 7 species were  In term of β diversity, the results poster a high level of similarity between plots with practices and controls plots in both localities (Table 4).
It could be inferred from these results a high level of similarity for flora diversity, between plots with practices and control (Sorensen Index over than 90%).

Effect of best land management practices on the structure of woody vegetation
In the context of large-scale use of agroforestry practices, as it is in Dan Saga cluster, the study highlighted a significant effect of ANTR practice on tree density (95 trees/ha on plots with practice against 38 for control plots). This effect is considerable on basal area and regeneration. The results of this study confirm those of several authors, including [17] who found a higher density of 141 trees/ha. The strong regeneration of woody species in the fields suggests a woody population dynamic characterized by a scarcity of seed companies. This is well illustrated by the vegetation structure, presented through the distribution of tree density according to diameter classes. Indeed, this structure displays an inverted J trend with a shape parameter quite close to 1. The abundance of juvenile subjects associated with this structure evokes fairly well the strong regeneration, especially in the case of plots with practices. In this sense [11] had shown that the regeneration of the dominant species (Annona senegalensis, Combretum glutinosum and Piliostigma reticulatum) is clearly very high than that of the other species. In the case of plots without practice, the inverted J shape of the curve is not strictly uniform, suggesting a structure with an uneven-aged tendency. In this case, it should be emphasized with [21] that despite the importance of regeneration, the high value of shape parameter may present some risks of survival during the development transition phases.
In the case of Tabofatt, where revegetation is linked to tree planting operations coupled with water and runoff collection structures, the effect of best practices on the structural parameters of the woody population is quite temperate. It is more noticeable on regeneration (regeneration rate of the plots with practices is 67% against 42% on the control plots). The results even revealed that the density of trees in the control plots is slightly higher than that of the plots with practice. Overall the tree density is a little less than 100 trees per ha. This means that the vegetation is predominantly shrubby as evidenced by the regeneration rate of over 50%. In South of Dan Saga near the Dan Kada Dodo forest, [13] had recorded an adult density of 25.2 plants per ha, significantly lower than the results of this study. This could be linked to the strong pressure on forest resources in that zone compare to the North one. In addition, the basal area in plots with practice was in both cases clearly greater than that of the control plots. This suggests that practices had a positive effect on landscape revegetation. Indeed, the result of basal area found by this study is of the same order of magnitude as those found by [27] in the Zakouma forest, and the one found by [28] in the agroforestry parks bordering the Tamou Total Wildlife Reserve in Niger. The population renewal rate recorded by this study fluctuates between 25 and 30%. This confirms the efforts of farmers in landscape restoration actions. For the case of plots with or without practice, the shape of the curve representing the diametrical structure of the vegetation looks like a bell. The value of the shape parameter is quite close to 2. That could be clearly linked to the influence of the revegetation of the landscape set by tree planting. Plants belonging to the 5-10 cm class are the most abundant. Therefore this distribution displays on the one hand the anthropic pressure on the spontaneous species and in other hand the regenerating driven mainly through seeds [18,29]. The importance of the local soil and climate factors which favors the species of the Mimosaceae family could also play a key role on the woody vegetation structure. In this sense, the surface states of the ground would probably have conditioned this stratification. In this regard, [30] highlighted the depressive effect of crusted surfaces on the characteristics of ligneous vegetation in a sub-arid zone of western Niger which is located at the same latitude as our study area. So, we can have floristically homogeneous beaches isolated by bare crusted surfaces which limit the effect of practices on the structural parameters and evolution of the woody population in the case of the Tabofatt cluster.

Effect of best practices on the composition and diversity of woody vegetation
The floristic procession expresses quite remarkable demographic characteristics depending on the context considered. In the Dan Saga site, plots with practices are dominated by Piliostigma reticulatum, Combretum glutinosum and Guiera senegalensis. [11] had shown that these three species constitute with Annona senegalensis the species dominating the landscape of Dan Saga. These are species characteristic of the practice of ANTR, with regard to their high potential for regeneration by rejection of stumps [17]; [31]. This dominance can be transposed to plots without practice. In overall, the plots with practice contain 50 species against 44 species for the control plots. In this context, the results of plots with practice confirm the positive effect of practices on floristic richness. Moreover, these results are of the same order of magnitude as those obtained by [12] who listed 54 species. On the other hand, they are clearly higher than those obtained by [32] who found 26 species and [11] who recorded 37 species, suggesting a gradual trend in floristic richness over time. At the Tabofatt site, the plots with practice are characterized by 22 species against 20 species for the control plots. These results are similar to those obtained by [14] who had identified 22 species in part of the Maggia basin. In summary, this shows that the effect of tree planting coupled half-moon or stone lines structures is not significant on the specific richness of the woody population in the case of Tabofatt. In terms of specific dominance, the results also show a slight difference between the plots with practice and the control plots. This may be due to the floristic simplification, since in this case the invasive Prosopis juliflora species in the riparian vegetation of the Maggia koris is strongly dominant (IVI = 48.59%). Obviously, this invasive species is decisive in the expression of regreening in the context of a relatively contrasting sandstone plateau. Indeed, several authors including [28,29,33] argued for this simplification by the influence of the anthropogenic factor either by selective harvesting or by selection and protection of young plants to the detriment of senescent one.
At plot level, the specific richness of plots with practice is significantly higher than that of plots without practice. Just like the value of the Shannon-Weaver diversity index which is 2.93 bits. This level of diversity is of the same order of magnitude as that obtained by [17] who found 3.48 bits and [11] who reported 2.2 bits. In fact, [31] had reported that in Dan Saga, a strict Sahelian zone, the practice of ANTR has made it possible to enhance and maintain a floristic diversity similar to that of the southern part of Aguié, a much wetter Sahelo-Sudanian zone. In this case the importance of the floristic richness is not linked to the rainfall gradient. Without denying the methodological disparities, these results indicate that there is in both sites a transformation of the ligneous cover which takes place jointly through the abundance of vegetation and the reinforcement of phytodiversity, under the influence of the use of best land management practices.

Conclusion
The recent re-greening of certain Sahelian areas is based on the large-scale use of the sustainable best land management practices. Therefore it appeared necessary to assess the effect of these practices on the structure, composition and diversity of the vegetation in these areas. This study highlighted the positive effect of the use of best land use practices on several parameters such as: regeneration rate, flora structure composition and diversity in agroforestry parklands landscape. The results shown that species such as Combretum glutinosum, Guira senegalensis and Piliotigma reticulatum were dominant in Dan Saga site while Acacia seyal and Prosopis juliflora were quite dominant in Tabofatt site. These two different vegetation facies could be explain by three factors namely soils characteristics, climate condition and types of best land management practices adopted in the site. In terms of soils characteristics Dan Saga is located in deph and sandy plateau while Tabofatt site is on and sandstone plateau with thin soils and quite deep cleyley soils in the valleys at. The climate condition has also played a key role with semi-arid at Dan Saga and sub-arid at Tabofatt site. Meanwhile the type of practices appeared as the driving factor of vegetation dynamic of the two landscapes. On one hand; the Assisted Natural Tree Regeneration practice widely used in Dan Saga site has induced the increase of young trees density (high regeneration rate) and flora diversity dominated by Combretaceae family. On the other side the planting of trees coupled with anti-erosion structures such as half-moon and stone lines are mostly used in Tabofatt site. That practice increased trees density within the sandy valley and along the streams banks as well as the increase of species richness dominated by Mimosaceae family. In the overall the vegetation pattern of the two landscapes displayed shrubs structure. These results constitute benchmarks that can guide the scaling up of re-greening promotion centered on the dissemination of best practices for sustainable land management. In perspective, further understandings of re-greening process in sahelian region require more studies based on remote sensing technologies coupled with plot monitoring of woody vegetation.