Globally, mangroves cover an area of 137000 km² [1]. These tropical coastal forest ecosystems are known for their high biomass productivity and carbon sequestration capacity [2,3]. A minimally disturbed mangrove sequesters on average 1087±584 Mg C/ha [4]. The conservation of mangroves appears to be an essential component for both the well-being to local human communities and the fight against global warming [5,6].

Despite their importance, mangroves are exposed to both natural and anthropogenic pressures that have resulted in the reduction of vegetation cover [7,8]. Overall, the rate of deforestation of mangroves was estimated at 4.3% in 1996-2016 period [1]. In total, nearly 1646 km² of the world's mangrove area disappeared between 2000-2012. Cameroon has 1113 km² of mangroves along its western Atlantic coast [9]. These mangroves have been subject to increased anthropisation over the last two decades, which is manifested by a considerable decrease of vegetation cover [10]. Mangroves of Douala-Edea National Park (DENP) are not spared from this. The annual rate of regression of these mangroves has been estimated at 53 ha [11]. These mangroves have been the main source of firewood used for smoking fish in various fishing camps and surrounding households for at least three decades [12].

In response to this ecological disaster, many options for mangrove conservation have been implemented around the world, including restoration. Mangrove restoration refers to a set of human actions aimed at re-establishing ecological processes, which accelerate the recovery of levels of forest structure, ecological functioning and biodiversity to those typical of climax forest [13]. More than 190147 km² of mangroves were restored worldwide by 2018 [14]. In Central Africa, more than 500 ha of degraded mangroves were reforested before 2017 [15]. In Cameroon, reforestation of degraded mangroves by governmental organisations and other partners officially started in 2009 and continues up to these days with the implementation of the planting of 40000 seedlings in the mangroves of Bolondo fishing camp village located in DENP [16].

Mangrove restoration faces many obstacles worldwide that result in complete or partial failures of restoration projects [17, 14, 18]. Not only do these failures cause ecological loss, but they also constitute a considerable financial loss as the average cost of restoring one hectare in Central Africa is estimated to US$ 3200 [15]. The lack of a large database of monitoring and evaluation of restoration projects was cited among the main causes of the barriers of mangrove restoration [19].

Along the African Atlantic coast, very few studies provide information on structure and functioning of restored mangrove areas [20]. In Cameroon according to our knowledge, only specific study on mangrove restoration is limited to the determination of main abiotic factors influencing the growth of seedlings in nurseries [21]. No published scientific study focusing on the monitoring and assessment of standing biomass and litterfall production of reforested mangrove areas of the DENP has yet been known. However, information from quantifying the biomass of reforested areas would be crucial for safeguarding mangroves of the DENP and for improving climate change mitigation strategies [22]. Moreover, it is important to integrate in any restoration project the monitoring of objectively verifiable indicators of ecosystem functioning such as litterfall production [23]. To contribute improvement of mangrove restoration projects in DENP, this study aims to assess the impact of mangrove reforestation through the estimation of both standing biomass carbon stock and litterfall production of three reforested mangrove areas with reference to their respective adjacent good natural stands.

Study Site

The study site is located in the mangroves areas of the Douala-Edea National Park (DENP) included between latitude 3°14′ N-3°50′ N and longitude 9°34′ E-10°03′ E. Mangroves occupy 39202.2 ha of the DENP or 14.91 % of its surface and belong to mangroves of the Cameroon Estuary (Figure 1).

Figure 1. Location of studied sites (Modified from [25]).

`The work was carried out in the mangroves areas adjacent to two fishing camps (Bolondo and Yoyo II) in Mouanko district (Figure 1). The climate is equatorial belonging to the Cameroonian type domain. It is marked by a short dry season (December to February) and a long rainy season (March to November which has a peak rainfall during September) with annual average temperature of about 26.7°C [24]. The tide is semi-diurnal. Five characteristic mangrove species (Avicennia germinans (Linn.) Stearn, Nypa fruticans (Thurnb.) Wurmb, Rhizophora harrisonii Leechman, Rhizophora mangle Linn. and Rhizophora racemosa Meyer) are abundant, associated with two other species (Ficus sp. and Raphia sp.) [12].

Data Collection

Delimitation of Permanent Sample Plots (Psp) and Measurements: Fifteen (15) Permanent Sample Plots (PSPs) of 100 m² (10 m × 10 m) were set up in three Reforested Mangrove Areas (RMAs), 5 in Bolondo and 10 in Yoyo II. In adjacent good natural mangrove stands of each reforested area, 5 PSPs were set up to serve as reference sites. The choice of these reference sites was inspired by the work of [26, 27, 28, 29] done at Gazi bay of Kenya, Jaguaribe River in Brazil and Perancak River in Indonesia respectively. A total of 30 PSPs were set up, 15 of which were in three RMAs and 15 in their respective adjacent good natural vegetation (Table 1). All the works conducted on the field were during low tide.

Table 1. Main information about PSP and litter traps.

Height was determined either using of a graduated stick or by using a Suunto PM-5360-degree clinometer. Diameter at breast height (DBH) were measured using a Vinier calliper for young trees and a forestry diameter tape for older trees.

Litterfall Collection: Litterfall refers of all organic debris from plants, decomposing, on the surface of the ground. To estimate litterfall production, five litter traps of 1 m² (1m×1m) were installed in each PSP [29]. These litter traps were made using polyester sieves with a pore size of 0.5 mm and they were set up to collect as much litterfall as possible from each of the PSPs. Two installation methods were used in the field depending on the average height of the individuals: (1) for the 5 PSPs in which the height of individuals was less than 3 m, traps were suspended by stakes at an average height of 1.5 m from the ground (Figure 2A); (2) for the other 25 PSPs in which the height of individuals was more than 3 m, the five litter traps were suspended on branches of the trees (Figure 2B).

Figure 2: Installation of litter traps in PSPs: A. for individuals less than 3 m high (traps indicated by arrows); B. for individuals more than 3 m high.

Collection was carried out every 30 days for 12 months (October 2020 to September 2021). The collected biological samples were stored in labelled plastic zip bags and taken to laboratory (Figure 3A). The samples were repackaged in similarly labelled A4 paper envelopes and oven-dried at 70°C to a constant mass (Figure 3B). The average drying time was 72 hours. The dried plant parts were sorted into flowers, leaves, twigs and seed-fruits, and weighed to the nearest 0.01 g using a Zhi Heng Digital Jewelry Scale, model ZH-8256 (Figures 3C and 3D).

Figure 3: Litterfall treatments: A. Collection of samples in the field, B. Oven drying, C. Twigs weighing, D. Flowers weighing.

Data Analysis

Vegetation Structure: Data from both plant species inventories carried out and measurements taken on individuals, were used to calculate the following vegetation study indices in accordance with [30]: Abundance (A), Densities (D); basal area of woody species i (STi) ; average diameter (dm); average height (Hm), Complexity Index (CI).

In order to compare the structure of studied vegetation, the diameters and heights of individuals were divided into 5 classes. The diameter classes were (in cm): ?3, [3-5[, [5-7[, [7-10[ and ?10 while the height classes were (in m): ?3, [3-5[, [5-7[, [7-10[ and ?10.

Biomasses and Carbon Stocks

Estimation of Standing Biomass and Carbon Stock of Reforested Mangrove Areas: Three biomass compartments were estimated in Permanent Sampling Plots (PSPs): (1) above-ground biomass (AGB); (2) below-ground biomass (BGB) and (3) litterfall production (see section 1.2.2). The standing biomass value corresponded to the sum of above-ground and below-ground biomass [31].

Allometric Equations

AGB and BGB rates of each individual PSP were estimated using the following specific allometric equations:

- AGB [32]:

• germinans: AGB = 0.14(DBH)2.4;
• Rizhophora spp: AGB = 0.1282(DBH)2.6;
• BGB [33]:
• General equation: BGB = 0.199×ρ0.899(DBH2.22). ρ: wood density of considered species. For A. germinans: ρA: 0.661 g/cm3; for Rhizophora spp.: ρR = 0.883 g/cm3.

These allometric equations were chosen because diameter ranges of the individuals sampled in reforested areas are similar to those of the individuals used for their design. The biomasses estimated from PSP were extrapolated to the hectare. Above-ground, below-ground and litter fall carbon stocks were estimated by multiplying AGB, BGB and litter fall production by their respective carbon concentration coefficients: 0.5; 0.39 and 0.45 [30].

Statistical Analysis

Statistical tests (ANOVA, t-test) were carried out to compare the values obtained for different parameters between reforested mangrove areas (RMAs) and their respective adjacent good natural vegetation. The "r commander" package of R software version 4.1.3 was used for descriptive statistics and to perform significance tests. Excel 2013 was used to produce histograms and biomass variation curves. RMAs of different ages and their respective adjacent good natural vegetation were considered as treatments and their PSPs were considered as replicates.

Statut of Vegetation of Rma

Composition Plant Species and Density: A total of three species belonging to two genera and grouped in two families were identified in sampled areas. Two species of the Rhizophoraceae plant family were sampled in Reforested Mangrove Areas (RMAs): Rhizophora mangle L. and Rhizophora racemosa Meyer, while those identified in the adjacent good natural vegetation were Avicennia germinans (L.) Stern (Acanthaceae) and R. racemosa. With exception of Bol-nat1, all sampled areas were monospecific. The abundances of RMAs were 4000±200 ind./ha, 3280±238.74 ind./ha and 2160±240.83 ind./ha for the 3-years, 6-years and 11-years mangrove stands respectively. These abundances showed a negative trend regarding ages of vegetation, mean diameters, mean heights and complexity indices (Table 2). This trend could be explained by progressive increase in competition between individuals for environmental resources, which would lead to a reduction of woody individual abundance over time. Like our observations, the abundance of RMAs decreases with the age of vegetation in the Gazi Bay in Kenya and in the Rufiji Delta in Tanzania [34, 35].

Distribution of Diameters and Heights

The mean diameters of individuals were 1.20±0.5 cm; 2.58±0.85 cm and 2.93±1.4 cm respectively for RMAs aged 3 years (Yoy-Plant2), 6 years (Bol-Plant1) and 11 years (Yoy-Plant1) (Table 2). The Student's statistical test showed no significant difference between the diameters of individuals in the RMAs Bol-Plant1 and Yoy-Plant1 (Table 2). The majority of individuals of RMAs were concentrated in ?3 cm class (Yoy-Plant2 (100%), Bol-Plant1 (63.63%) and Yoy-Plant1 (59.09%)) (Figure 4). Concerning the higher extremity, only individuals from adjacent good natural stands had representatives in last two intervals.

Figure 4: Distribution of Diameter classes.

The average heights of individuals were 1.8 (±0.3) m; 5.64 (±1.87) m and 3.34 (±1.26) m respectively for Yoy-Plant2, Bol-Plant1 and Yoy-Plant1 (Table 2).

Table 2: Specific composition and structure parameters of the vegetation of the sampled areas.

The Student's t test did not show any significant difference between the heights of the individuals of the RMAs Bol-Plant1 and Yoy-Plant1. The distribution of height classes shows that the individuals of RMAs had heights below 10 m (Figure 5).

Figure 5: Distribution of height classes.

In general, the average heights of individuals from the respective adjacent good natural stands were higher than each of their RMAs (Table 2). This could be explained by the fact that the respective adjacent vegetation is older than the reforested vegetation. The average height of individuals in the 11-year-old RMA (Yoy-Plant1: 3.34±1.26 m) is lower than that of the 6-year-old (Bol-Plant1: 5.64±1.87 m). This result is contrary to previous ones and could be explained by the difference of plant species used for reforestation at the two sites. The mean height and mean diameter of individuals in the 3-year-old RMA are similar to results obtained in a 3-year-old R. racemosa RMA in Kono Creek, Nigeria [20].

There is a correlation between increase in height and diameter of individuals of the same species and this is a function of the age of RMA. This finding is similar to that for seedling growth in Perancak estuary in Bali, Indonesia [29]. In the same RMA of Rhizophora mucronata Lam. in Kenya, the mean heights of individuals were 4.70±0.20 m and 8.40±1.10 m at 8 and 11 years respectively [26, 34]. Complexity index values did not vary proportionally with the age of vegetation in the reforested areas. However, those of the reforested areas (Yoy-Plant2 (0.40×10-3), Bol-Plant1 (3.54×10-3) and Yoy-Plant1 (1.14×10-3)) were lower than index values of their respective adjacent good natural stands (Yoy-Nat2 (76.1×10-3), Bol-Nat1 (390×10-3) and Yoy-Nat1 (32.95×10-3)).

Standing Biomass and Carbon Stock

Above-Ground Biomass (Agb) And Below-Ground Biomass (Bgb): The above-ground biomasses (AGB) of RMAs did not vary strictly with the age of the vegetation although the smallest value was for the youngest RMA (Table 3). The one-tailed Student's t test with the same variance showed no significant difference between AGB and below-ground biomasses (BGB) of 6-year-old and 11-year-old RMAs (p? 0.05).

The respective AGB and BGB of adjacent good natural stands (Yoy-Nat2 (2653.13±608.91 kg/ha and 1289.07±422.02 kg/ha), Bol-Nat1 (6857.94±563.07 kg/ha and 2907.77±1111.95 kg/ha) and Yoy-Nat1 (575.64±57.72 kg/ha and 362.78±73.72 kg/ha) were higher than those of the RMAs (Yoy-Plant2 (11.98±0.76 kg/ha and 13.88±1.3 kg/ha), Bol-Plant1 (61.18±2.16 kg/ha and 55.19±1.92 kg/ha) and Yoy-Plant1 (55.25±2.93 kg/ha and 47.8±1.31 kg/ha) (Table 3). Furthermore, statistical test showed a significant difference at the 5% threshold between the biomasses of the two vegetation types.

Table 3: Biomass and carbon stock of study sites.

The total standing biomass of RMAs of Douala Edea National Park (DENP) is positively correlated with age. This result can be explained by the difference in the diameter of individuals of the two types of vegetation sampled and the growth in stem diameter is positively correlated with the age of the vegetation. This observation is like that of [36] in Vietnam and [35] in Tanzania where [35] obtained total carbon biomass values of 13.65; 20.13 and 57.53 Mg C/ha for 5-, 10- and 15-year-old RMAs respectively. On the other hand, a biomass of 60.43 Mg/ha was estimated for a 5-year-old R. mangle RMA in Potenji estuary in Brazil [28]. In a 12-year-old R. mucronata RMA in the Gazi Bay, Kenya, above-ground and below-ground biomass values were 106.7±24 t/ha and 24.9±11.4 t/ha respectively [37]. These values are very high compared to that of the 11-year-old R. mangle RMA in DENP. This disproportionate difference in biomass between ages can be explained by the species used for reforestation on the one hand, and by the evaluation method used on the other hand [34]. The higher biomass of the natural vegetation of RMA in DENP would be an indicator of the fact that RMAs have not yet fully restored the functional properties of the pre-existing ecosystem.

Biomass and Carbon Stock of Litterfall in the Sampled Areas

Litterfall Composition: In all sampled vegetation, litterfall was composed of: branches, leaves, flowers and seed fruits. Table 4 shows the annual distribution of different components in the sampled areas. It can be seen that the proportion of dry leaves biomass is higher in the sampled areas irrespective of stands type and age of RMA. For RMAs, the proportions of leaves were: 93.08% (Yoy-Plant2), 95.93% (Bol-Plant1), 80.61% (Yoy-Plant1) while those of the respective natural vegetation were: 67.84% (Yoy-Nat2), 82.43% (Bol-Nat1), 90.97% (Yoy-Nat1) (Figure 6). In contrast, the proportion of dry fruits-seeds biomass was zero only for the 3-year-old RMAs. In this and many other studies related to litterfall production from mangroves, the proportion of eaves was dominant [38, 27, 39, 40, 29].

Figure 6: Distribution of the proportions of annual productions of litter fall components by site.

Total Annual Litter fall Production: Dry litterfall biomasses varied in the same direction as the age of the RMAs: Yoy-Plant3 (40.78±7.42 g/m²/year), Bol-Plant1 (397.75±75.79 g/m²/year), Yoy-Plant1 (576.23 ±106.75 g/m²/year). The litterfall biomasses of RMAs were significantly lower than those of their respective natural vegetation: Yoy-Nat2 (929.48±203.38 g/m²/year), Bol-Nat1 (722.93±28.34 g/m²/year), Yoy-Nat1 (719.31±70.89 g/m²/year). Furthermore, Tukey HSD multiple comparison tests between the average of monthly dry litterfall production between the different pairs of RMAs and their respective adjacent good natural stands showed a significant difference.

Table 4: Components and production of litterfall in three reforested areas of different ages and in their respective adjacent vegetation.

Litterfall production in mature mangroves worldwide has been estimated to be between 2 and 16 t/ha/year [38]. Those obtained in this work (Yoy-Plant2 (0.4 t/ha/year), Bol-Plant1 (3.97 t/ha/year), Yoy-Plant1 (5.76 t/ha/year)) belonged to this range except for the 3 year-old RMA (Yoy-plant2). These litterfall productions show that these two ecosystems would contribute to the restoration of ecosystem functionality [37]. Furthermore, the litterfall production of the 11-year-old RMA (Yoy-Plant1) showed no significant difference with its adjacent good natural stands. This result is similar to that observed in the R. mucronata RMA older than 11 years in the Gazy Bay. As litterfall production is an important ecological function of the vegetation, these two mangrove stations would function similarly in their respective ecosystems. This would imply that the productivity of the reforested areas would be able to match that of the natural stands from a certain age under favourable conditions [27].

 Monthly Litterfall Production and Factors of Varation: Similar to the annual production, monthly dry litterfall biomass varied with the age of the plantation. Figure 7A shows that on the one hand variation of the dry litterfall biomass is similar in the 3 and 6 years old RMAs throughout the year and on the other hand, the older RMA showed the higher dry litterfall biomass. Furthermore, there was a significant difference between the monthly dry biomass values for RMAs of different ages (F = 38.86; P = 2.38291×10-09). In contrast, the dry biomass values for natural vegetation of RMAs respectively showed very similar variations over time and their dry biomass values were not significantly different at the 5% threshold (F= 2.49; P= 0.098) (Figure 7B).

Figure 7: Variation in monthly production of dry biomass of litterfall: A. In the three reforested areas of different ages; B. In the adjacent good natural stands.

The comparison of the mean dry litterfall biomass values between each of RMAs and their respective adjacent good natural stands shows that dry litterfall biomasses of the adjacent good natural stands were higher than those of RMAs (Figure 8). However, the mean monthly dry litterfall biomass values of 6-year-old RMA (Bol-Plant1) were similar to those of its adjacent good natural stands (Bol-Nat1) but the one-way Student's t test with different variances showed a significant difference between these two sets of values at the 5% threshold (t= -3.37; p = 0.0022). Furthermore, the one-way Student's t test with equal variances showed that there was no significant difference between the monthly biomass averages of the two areas Yoy-Plant1 (11 years) and Yoy-Nat1 (t= -1.516; p = 0.0716).

Figure 8: Comparison of litterfall dry mass between reforested mangrove and their respective adjacent good natural stands.

The superposition of the monthly variations of dry biomass presented in Figure 8 and the annual variation of the seasons in the locality show that dry biomass of litterfall decreases with increasing rainfall. Furthermore, maximum litterfall production was recorded in the dry season (November 2020 to February 2021) while minimum production was recorded during the months of heavy rainfall (June 2021 to August 2021) with the exception of Yoy-Nat1 PSP in which there was a significant fall in twigs during June 2021 (Figure 9A).

Thus litterfall production of the RMAs of the DENP varied with both age and season. This result is similar to that observed in several mangroves: the Perancak mangroves in Bali, Indonesia, Sundarbans in India and Zanzibar mangroves in Tanzania [39, 29]. In the RMAs of DENP, litterfall production increases with decreasing rainfall. This could be explained by the fact that there is a reduction in freshwater supply in estuary which is accompanied by an increase in the salinity of the environment and the plants adapt to this saline stress by losing leaves and some branches to reduce water loss [38].

The dry twig biomass collected was discontinuous and very low throughout the year for the 3 and 6-year-old RMAs (Figure 9A). In contrast, twigs biomass in the 11-year-old RMA was higher than in the other two RMAs, and continuous throughout the year. Moreover, its variation is very similar to that of its adjacent good natural stands. The annual variation in fruit/seed dry biomass showed the same characteristics as that of the twigs except that the months of maximum and minimum biomass collection were different (Figures 9A and 9D).

The dry biomass of collected leaves and flowers was continuous throughout the year in the respective adjacent good natural stands and in the 11-year-old RMA (Figures 9B and 9C). The biomass of flowers collected was almost zero throughout the year in the 3-year-old RMA, whereas the variation in the biomass of the 6-year-old RMA is very similar to the variation in its adjacent good natural stands (Figure 9C). In addition to being very dominant in litter fall composition for all PSPs, leaf biomass production was continuous in all PSPs throughout the year (Figure 9B). In general, litterfall production of all mangrove species worldwide is influenced by air temperature, insolation and rainfall, forest succession stages (i.e., pioneer plants-young forests-mature forests), forest management (e.g., selective pruning or harvesting) and anthropogenic disturbance (e.g., coastal development) [40, 29].

Figure 9: Monthly variation in dry biomass of litterfall components: A. Branches; B. Leaves; C. Flowers; D. Fruits-seeds.

The assessment of standing biomass and litterfall production of three reforested mangrove sites of DENP showed that Rhizophora was the plant genus used for reforestation in the sampled RMAs. Average diameter, average height, standing biomass and litter production were positively correlated with age of the reforested areas. With age, litter production could approach that of the adjacent natural vegetation. This result would be an ecological success indicator of the DENP mangrove restoration projects at the short term. Thus, reforestation of the disturbed mangrove of the DENP would make a significant contribution to the safeguard of these ecosystems. However, the allometric equations used here were developed on the basis of vegetation data that do not originate from African Atlantic coast. It would be necessary to develop original allometric equations to these ecosystems in order to improve accuracy of the estimate.

The authors would like to thank Cameroon Wildlife Conservation Society (CWCS) for its materials support that permitted to realize this work on the field.

1. Spalding M, Leal M. The State of the World’s Mangroves 2021 Global Mangrove Alliance, 2021.
2. Alongi DM. Carbon Payments for Mangrove Conservation: Ecosystem Constraints and Uncertainties of Sequestration Potential. Environ. Sci. Policy. 2011; 14: 462-470.
3. Chowdhury A, Naz A, Maiti SK. Variations in Soil Blue Carbon Sequestration between Natural Mangrove Metapopulations and a Mixed Mangrove Plantation: A Case Study from the World’s Largest Contiguous Mangrove Forest. Life. 2023; 13:1-11.
4. Sasmito SD, Sillanpaa M, Hayes MA, Bachri S, Saragi-Sasmito MF, et al. Mangrove Blue Carbon Stocks and Dynamics are Controlled by Hydrogeomorphic Settings and Land-Use Change. Glob. Chang. Biol. 2020; 1-6.
5. Din N, Ngo-Massou VM, Essome-Koum, et al. Local Perception of Climate Change and Adaptation in Mangrove Areas of the Cameroon Coast. J. Water Resour. Prot. 2016; 608-618.
6. Kathiresan K. Mangroves: Types and Importance. In Rastogi RP, Gupta DK, Phulwaria M, ed. Mangroves: Ecology, Biodiversity and Ecology, Springer Nature Singapore, 2021: 1-31.
7. Goldberg L, Lagomasino, D, Thomas, N. Global Declines in Human?driven Mangrove Loss. Glob. Chang. Biol. 2020; 26:1-32.
8. Emane JM, Essome-Koum GL, Ngotta-Biyon J, Ekodeck, GE, Tomedi-Eyango M, Din, N. Effects of Anthropogenic Pressures on the Structure of Floristic Components of Mangroves in the Cameroon Estuary. Examines Mar. Biol. Oceanogr. 2021; 4:1-7.
9. Hamilton SE, Casey D. Creation of a High Spatio-Temporal Resolution Global Database of Continuous Mangrove Forest Cover for the 21st Century (CGMFC-21). Glob. Ecol. Biogeogr. 2016; 25: 729-738.
10. Din N, Ngo-Massou VM, Essome-Koum GL, Ndema-Nsombo E, Kotte-Mapoko E, Nyamsi-Moussian L. Impact of Urbanization on the Evolution of Mangrove Ecosystems in the Wouri River Estuary (Douala Cameroon). Coastal Wetlands?: Alteration and Remediation. Springer, Coastal Research Library; 2017; 81-131.
11. Ajonina GN, Usongo L. Preliminary Quantitative Impact Assessment of Wood Extraction on the Mangroves of Douala-Edea Forest Reserve, Cameroon. Trop. Biodivers. 2001; 7: 137-149.
12. Ajonina GN. Inventory and Modelling Mangrove Forest Stand Dynamics Following Different Levels of Wood Exploitation Pressures in the Douala-Edea Atlantic Coast of Cameroon, Central Africa. Thesis in the Faculty’ of Forest and Environmental Sciences, Albert-Ludwigs-Universität, Germany; 2008; 215.
13. Elliott S, Blakesley D, Hardwick K. Restauration des Forets Tropicales: un guide pratique. Royal Botanic Garden, Kew. 2013.
14. Worthington T, Spalding M. Mangrove Restoration Potential: A Global Map Highlighting a Critical Opportunity. The Nature Conservancy. 2018; 34.
15. Ajonina GN, Aya FA, Diame A, Armah AK, et al. Overview of Experience of Mangrove Reforestation in West and Central Africa. In Ogunsanwo OY, Akinwole AO, eds., 38th Annual Conference of FAN. Forestry Association of Nigeria. 2016; 17: 12-21.
16. Planete-Urgence. Projet Cameroon Mangrove Ecosystem Restoration and Resilience. 2023.
17. Lopez-Portillo J, Lewis RRI, Shing VHR, Lee Y. Mangrove Ecosystems?: A Global Biogeographic Perspective Structure, Function, and Services. A Global Biogeographic Perspective: Structure, Function, and Services. 2017; 301-345.