Evaluation of the Aerial Biomass of Three Sahelian Species in the Ferlo (North Senegal): Acacia tortilis Acacia senegal (L.) Willd and Balanites aegyptiaca (L.)

In a context of climate change characterized by rising temperatures, increased greenhouse gases and frequent droughts, the Sahel is presented as one of the most vulnerable areas to the adverse effects of climate change. The Sahel is presented as one of the most vulnerable areas to the adverse effects of climate change. The objective of this study is to assess the above-ground biomass and carbon stock of three Sahelian species: Acacia raddiana tortilis (Forsk.) hayne ssp. raddiana (savi) Brenan, Acacia senegal (L.) Willd and Balanites aegyptiaca (L.). The study was carried out in northern Senegal commonly known as Ferlo. Biomasses of the populations of area is estimated at 1777.008 tons of CO 2 . The comparison between the models developed in this study and the multispecific or mono-specific models from the literature showed substantial differences. This study contributes to a better understanding of the contribution of Sahelian woody species to carbon sequestration and the results could be used in the framework of adaptation to climate change.


Introduction
From the drought years of the 1970s to the 1990s, the Sahel region underwent profound ecological and socio-economic changes [1] [2]. Recent studies on the evolution of Senegal's ecosystems in the Ferlo zone have shown a detrimental change in the floristic composition and structure (density, stratification and cover) of the vegetation [3] [4] [5].
In Sahelian ecosystems, the plant biomass produced decreases when the annual rainfall is less than 250 mm; even though soil conditions and human activities have a great influence [6]. This biomass provides important information on the functioning and economic productivity of agrosystems as well as their potential for sequestering atmospheric carbon [6].
Thus, under conditions of optimal management of Senegal's Sahelian ecosystems in a Great Green Wall (GGW) context, knowledge of its functioning and the services it provides is necessary in order to enhance its potential. According to the Intergovernmental Panel on Climate Change [7], significant terrestrial carbon stocks are vulnerable to the impacts of climate change and/or land-use change. To address these concerns, it is important to provide this international community with recent scientific information useful in the fight or mitigation of climate change due to the emission of greenhouse gases, especially carbon.
In the Ferlo area, relatively few studies have focused on estimating woody biomass. The few existing works are the subject of studies correlating aerial biomass and their dendrometric characteristics [8]- [14].
This work is approached at the population scale of three species reforested in the Ferlo in the area of the Great Green Wall; these are Acacia tortilis (Forsk.) hayne ssp. raddiana (Savi) Brenan, Acacia senegal (L.) Willd and Balanites aegyptiaca (L.) Del. It aims at taking stock of the carbon sequestration capacity of these woody species with a view to monitoring their dynamics with the evolution of carbon stocks as an indicator.
The selected species have strong economic potential for the local population.
In addition to their fodder role, these trees play an important role in the balance of the Sahelian ecosystem. They play a role in maintaining soil fertility and production [2] [3] [6], protecting the soil against bad weather (wind, run-off water,  [6]. Apart from their ecological and fodder importance, these three species contribute to the socio-economic resources of rural and urban populations. They are often used to satisfy needs in the fields of food (direct consumption of leaves and fruits), clothing (dyeing), energy (firewood), medicine (care of human and animal diseases), agriculture (tool handles, nitrogen fixation in the soil, restoration of depleted soils, regeneration of soils sterilized by salinity or alkalinity), and combating desertification (dune stabilization).
The estimation of the quantity of carbon sequestered by these woody trees is based on the principle of allometry according to which the relationship between certain parameters (height, diameter, crown size and biomass), obeys a rule that is the same for all trees living under the same conditions, from the smallest to the largest [15] [16]. The approach adopted in this study consists first of evaluating the biomass of woody trees by the destructive method and then of establishing some allometric equations allowing to predict the biomass of woody trees from the dendrometric characteristics and finally to deduce the quantity of sequestered carbon.

Presentation of the Study Area and Study Sites
This study was conducted in Senegal in the sylvopastoral zone commonly known as Ferlo (Figure 1). The Ferlo covers an area of 70,000 km 2 [17] and is crossed from east to west by the route of the Great Green Wall project in Senegal. Climatically, the annual rainfall of the Ferlo varies between 100 mm and 350 mm [1]. Relative air humidity is very low (annual average 35%) with high evapotranspiration ranging from 1800 to 2200 mm/year [1] [18]. The Ferlo has a woody to shrubby steppe dominated by woody trees such as Balanites aegyptiaca  This vegetation is inferred from two types of soils [20]: sandy iso humic soils of the dune system consisting of sandy red-brown clayey-sandy soils and poor in organic matter; and sandy to sandy red clayey-sandy tropical ferruginous soils, more or less leached and poor in organic matter.
The study was carried out in the western Ferlo center, in the environment circumscribed by five localities: Linguere, Dodji, Tsetsere, Widou Thiengoly and Kamb ( Figure 1). (P = percentage of the species; a = the ratio of the density of the species studied to the total density of woody plants; d = is the error rate; U = is the u of the normal law which is read from the table; n = number of feet to be sampled).

Wood Sampling
Based on this calculation, a total of 132 individuals were counted on the three transects ( Table 1).
The sample size in the elaboration of allometric models is variable, depending on the study areas, the density of trees and the spatio-temporal variability of precipitation [15]

Dendrometric Measurements
The dendrometric parameters measured on each individual were trunk diameter at 0.30 m and 1.30 m from the ground, and crown diameter on the N-S and E-W axes (Baskerville, 1965). Tree diameters were measured with a forest compass (Mantax black model A060). A tape measure was used for circumference measurements. For the feet of irregular and branched trunks, measurements were taken on the whole trunk; the individual values were then summed. The total heights of the individuals were measured on the ground after felling.

Felling, Cutting and Sampling of Trees
The selected individuals were cut down with a chainsaw. After felling, the aerial part of each individual was cut into pieces and divided into four categories or compartments: trunk, branches, twigs, and twigs and leaves [30]- [38].
To measure the dry biomass of the trees, branch samples were taken in all four directions (N, E, W and S); for twigs and leaves, one sample was taken in the northern direction only, since there was no significant difference in moisture content in the direction of sampling [39]. In this study, the underground biomass was not studied despite the fact that it constitutes a significant part of the Sahelian environment where the deep growth of trees would be very determining.

Weighing and Monitoring the Moisture Content of the Biomass in the Different Compartments
After slaughter, the total fresh mass of each compartment of an individual was weighed with a 200 kg ± 0.1 scale and then added to obtain the total fresh mass epigraphed. Samples from each compartment were taken and weighed more accurately with a precision balance (7 kg ± 0.05) and reported to the laboratory. At the laboratory, the samples were oven dried: leaves at 70˚C and wood at 105˚C until constant mass. To measure the moisture content of the trunks, washers at 0.30 m and 1.30 m from the ground were used.

Data Processing
The ratio of fresh and dry masses made it possible to calculate the moisture content of the different compartments of each individual. The following formula was used: Knowing the moisture content (Th), the dry biomass of the different compartments was calculated by the following formula: The calculation of the amount of carbon (C) stored by each species was determined from the dry biomass (BS). It is also based on the work of [40] which considers that 47% of the dry biomass of woody species is made up of carbon.
( ) Above-ground carbon stock (ton/ha) of woody species in the Ferlo area was assessed using the average carbon stock of the three species, but also based on the actual density of Ferlo (108.08 ± 49.79 ind/ha) according to [41]. The estimation of the carbon equivalent, atmospheric CO 2 , of the area was carried out by multiplying the volume of carbon by 3.67 according to the method used by [42] according A comparison of generic and mono-specific models from the literature with the models constructed in this study was then carried out. They are used to determine the relationship between biomass and wood scaling [14] [43] [44]. This is the model integrating the power function and the model integrating the polynomial function (from degree two to three).

Relationship between Dry Biomass and Dendrometric Parameters
In this study, the dry biomass of each of the three species is correlated to three dendrometric parameters: diameter at breast height (DBH), total tree height and crown diameter. Thus thanks to the selection and validation criteria, the best models are selected ( Table 2).
The analysis in Table 2 shows that the selected models met the conditions of  normality, independence, homoscedasticity and null mean of the residuals (p-value greater than 0.05). In addition, the estimated coefficients of the models are highly significant, which was translated into an overall significance (p < 0.001). The

Comparison of the Results of This Study with Models in the Literature
The results of this study are compared on the one hand with generic models [12] [24] [44] and on the other hand with mono-specific models developed by [8] [45] ( Table 3). The presentation of its allometric models in graphical form visually illustrates the degree of similarity between the models proposed in the literature and those from this study (Figures 3-6).
Comparison between the generic models and the model developed in this study shows that the former overestimate woody biomass ( Figure 2). However, it can be noted that the model in [12] is closer to the model developed in this study, followed by the model in [45]. With the model of [25], the predicted biomass increases very rapidly when the diameter reaches 7 cm. At 15 cm diameter, the biomass is largely overestimated compared to that predicted by the other models.
For the monospecific models, the model of [8] also overestimates the biomass of A. senegal ( Figure 6) compared to the model developed in this study. Only       Table 3. Generic models and mono-specific models from the specialized literature with the models developed in this study.

Assessment of Biomass and Carbon Stock by the Destructive Method
The biomass obtained by the destructive method after drying the samples is 31

Model Developed in This Study
The biomass prediction models developed in this study were used to calculate the biomass of the three species. The results obtained show a total biomass of

Existing Generic Models
Biomass predictions from the generic models compared to the destructive method (2,370,709 kg); show values twice as important for the model of [45] and [12] with respectively 5818/844 kg and 4,510,332 kg; and four times as important for the model of [25], i.e. 8,658,737 kg. For these comparisons, the Kruskall Wallis test showed highly significant differences (p-values less than 0.001) ( Table 6).

Existing Single-Species Models
Biomasses obtained by prediction with monospecific models appear to be higher than those obtained by the destructive method (Table 7).

Discussion
The allometric relationships that were the subject of this study were compared to existing models. The three (3) allometric models developed in this study each have an adjusted R 2 coefficient of determination close to 1 and meet the various preliminary statistical tests. The CSR values ranged from 0.10 to 0.13 as did the AIC values, which ranged from 10,223 to 15,423. These criteria are taken individually or combined in the choice of the best models by many authors [43] [44] [47]. For the latter, the best model is the one with the lowest AIC and CSR values. The same approach has been used by [48], who selected cubage tariff models by combining these two parameters. On the other hand, [49] developed allometric equations whose selection was based mainly on the low value of CSR. Thus, biomass prediction from these three models produced results very close to those of the destructive method.
The highly significant differences between the mass obtained by the destructive method and that obtained from the generic models developed by [12] [25] [45] indicate that it would not be appropriate to use these models in the context of Ferlo. Rather, his models tend to overestimate dry biomass.
Monospecific prediction models seem to better predict the biomass of the species studied, notably the model of [8]  According to [12], the use of a generic model amounts to introducing a prediction bias that can be seen as interspecies variability. His study shows that merging data for several species is of interest if the gain in intra-species variability brought by this merger compensates for the interspecies variability introduced. However, it is necessary to ensure that: 1) this merging makes sense and 2) the proportions within the sample of the different species present in the stand are respected. When constructing an all-species tariff from the outset (as is often of individuals in the sample is independent of their species, so as not to bias the tariff in favour of a particular species [12]. However, the best strategy to achieve the tariff for a stand is to construct monospecific allometric equations for the main species present; and then to carry out the interspecific assessment of dry biomass by applying the monospecific allometric equations while respecting the characteristics of the stand under study (species composition and proportion, but also distribution of tree diameters and sizes). By merging the data, the sample size is increased, which is of interest if it compensates for the increase in variability due to the mixing of different species [12]; but this is a rather cumbersome task to carry out.  [12] in the Sudanian and Sudano-Guinean savannas of Senegal (classified forests of Bala, Kantora, Mampaye, Ouli, Patako and Welor) where the biomasses obtained with the polynomial allometric model vary from 7.9 to 102.2 tMS/ha for equivalent carbon stocks from 3.93 to 50.89 tC/ha. In the wetter regions, the amounts of sequestered carbons are between 81.48 and 118.36 tC/ha [60]. These differences would be related to climatic conditions that influence in one way or another the size of the woody plants. [12] had pointed out that the value of biomass and carbon in the savannah is low because of the predominance of small diameter woody trees. Also, logging, which most often targets large trees, drastically reduces biomass stocks over the years.
Highly vulnerable to climate change, the Sahel, and more particularly the Ferlo, nevertheless presents significant assets for reducing the concentration of greenhouse gases in the atmosphere. Firstly, it contains vast areas still available for forest management and favourable to the development of adapted species. Secondly, reforestation combines very well with the fight against poverty, since forest resources weigh heavily on the economy of the population. Finally, the species best adapted to semi-arid climates are also those that provide the highest added value products, such as gum trees [2] [3]. On these bases, the Sahel is a zone of great potential interest in the perspective of promoting carbon sequestration while fostering new sources of income to fight against poverty and improve food security for the populations of the zone. Thus, estimating the quantity of carbon sequestered in Sahelian ecosystems as accurately as possible is of great interest for the development of the zone and its better inclusion in the global balance. The limits of this study are related to the destructive method used, but also the number of species considered in relation to the important woody diversity in the Ferlo area. However, it is essential to refine a new non-destructive methodology based on major species of the area for a better estimation of the carbon stock.

Conclusions
This present study carried out in the Ferlo on populations of A. raddiana, A. senegal and B. aegyptiaca allowed building on the one hand, mono-specific allometric models linking dendrometric parameters to the aerial phytomass and on the other hand, evaluating the quantity of carbon sequestered by these species in the Ferlo.
The study confirms previous research on the measurement discrepancies found between mono-specific and generic models established for the estimation of biomass and carbon stock. The woody species of the Senegalese Sahel, due to their small trunk size, have a low carbon sequestration capacity compared to other more humid ecosystems. Differences in the biomass prediction results can be explained by the variability related to ecological conditions, the choice of parameters to be considered in developing the model, and the choice of the mathematical model that best fits the experimental point cloud.
Estimating biomass and quantifying the carbon sequestered by woody plants in the Ferlo will gain in precision and reliability with single-species models, which take more account of the specificities of each species present in its stand.
The best strategy to achieve stand pricing would therefore be to construct monospecific allometric equations for the main species present; and then to carry out interspecific assessment of dry biomass by applying the monospecific allometric equations.
This study contributes to the evaluation of the contribution of Sahelian woody species to carbon sequestration in a context of climate change and to the monitoring of their dynamics with the evolution of woody carbon stocks as an indicator.