Managing a Reconstituted Rainforest Remnant and its Threatened Plants in Benin

Abstract

Background

In Benin, in the savannah of the Dahomey Gap, sacred forests of <0.5 ha. preserve the rainforest flora of the adjacent forest blocks ‘Upper Guinea’ and ‘Congo Basin’.

Research Aims

To protect these threatened plants, we reconstituted a rainforest patch of 14 ha, the ‘Sanctuaire des Singes’, out of fallow land and followed its regeneration during 28 years under threatening urbanization and climate change.

Methods

The growth of saplings of three common tree genera and the changes in the composition of the flora developing from seeds and roots were analyzed. Species collected in other rainforest patches were introduced. Management consisted in hand irrigation and mulching of young plants during dry periods and the removal of strangling climbers. The first author’s residency in the forest and support to the local, mostly poor (by UN standards) population with its vodun culture is highlighted.

Results

The forest, under IITA property, harbors 590 plant species; another 64, mostly from drier origins, disappeared during the study. A total of 257 species were introduced; 58 are threatened according to IUCN criteria, twelve of them critically endangered. Most species (56.6%) have only <4 specimens. Among common trees, about 1% died each year, while common herbs and woody plants of pan-African or pantropical origin, which thrive in the vicinity of the forest, disappeared from the closed forest.

Conclusion

This is the first reconstituted forest in Benin, a ‘reference forest’.

Implications for Conservation

The ‘Sanctuaire’ is of high educational and scientific value. Because of its easy access and despite increasing urbanization, it receives numerous visitors (mostly attracted by the critically endangered red-bellied monkey) - a timid start for ecotourism. The forest serves as an example for better managing sacred forests in the nearby Ouémé floodplain, a UNESCO World Heritage site; but increasing storm damage threatens its survival.

Introduction

At the end of the ice ages some 10 000 years ago, the surviving tertiary rainforest refuges in today’s Côte d’Ivoire and Cameroon started expanding through the dry savanna of West Africa, forming the current forest blocks with their typical flora of so-called Upper Guinea and Congo Basin origin, respectively. The forest never closed over the so-called Dahomey Gap of today’s Benin, Togo and southern Ghana, where the savannah reaches the Atlantic Ocean. Nevertheless, tiny forest pockets developed; they became the about 1000 sacred forests, located mainly in the Ouémé floodplain in southern Benin (Poorter et al., 2004; Giresse, 2008).

In a highly populated region with 250 people per km² (INSAE, 2013) and an agricultural landscape embedded in human-induced, so-called derived savannah (Mama et al., 2014; Paradis & Houngnon, 1977), sacred forests are islands of high biodiversity. Rainforests cover only 2% of the national territory, but harbor 20% of all plant species and 64% of threatened plants, according to IUCN criteria (Adomou et al., 2011). These islands of biodiversity lay mostly outside established nature reserves, making their protection the highest priority for nature conservation in Benin (Adomou, 2005; Sinsin & Kampmann, 2010; Neuenschwander et al., 2011). Apart from a detailed study by Adomou (2005), the flora and vegetation of sacred forests has mainly been studied during short surveys with correspondingly short species lists (Adjanohoun et al., 1989; CERF Bénin, 2013; Hèdégbètan, 2011; Julé-Beaulaton, 2008; Kokou et al., 2008; Nagel et al., 2004; Sokpon & Agbo, 1999) and not followed up over many years, as is the case in this study. Many sacred forests are seriously degraded, calling for rehabilitation measures.

Here we describe a 28-year effort to link up and rehabilitate forest fragments, situated at the outer edge of the growing population centers of Cotonou and Abomey-Calavi, by encouraging regrowth and by introducing lost and rare species, collected in southern Benin. The management of the forest, its flora, experiences with its most important inhabitants, two groups of the critically endangered, endemic red-bellied monkey, Cercopithecus erythrogaster erythrogaster (Cercopithecidae, Primates) (Neuenschwander, 2024), the history of the forest, the interactions with the villagers and the local vodun practitioners (Bello Bravo, 2000; Neuenschwander & Adomou, 2017; Neuenschwander et al., 2015) are updated here.

Seven years after the last assessment of the flora (Neuenschwander & Adomou, 2017), awareness of the urgency to protect the sacred forests of the Ouémé floodplain, which has meanwhile been assigned the status of a UNESCO World Heritage site, is growing. The threatening urbanization calls for a reevaluation of the observed transformative changes achieved by the rehabilitation strategy of introducing rare native species at a certain stage (secondary forest) of the natural forest regrowth. Being the only rehabilitated forest in Benin, the ‘Sanctuaire des Singes’ (monkey sanctuary), recently called a ‘reference forest’ (B. Sinsin, pers. comm.), is at the center of an effort to gain government protection of sacred forests beyond the existing legislation (Republic of Benin, 2012).

The present study documents 1- successional changes and species composition following the creation and sustainable management of a sanctuary for threatened flora. This natural forest also protects rare animals and serves as an example for the protection of other forests. 2- It describes the human environment and the conditions needed for the local population to accept and thereby protect the forest, as well as its hazards and benefits. 3- It describes the challenges posed by population growth and the ensuing urbanisation as well as the effects of climate change.

Materials and Methods

Study Site

Climate

The Dahomey Gap has a strong gradient of rainfall with the highest values towards the East, i.e., the Nigerian border (Le Barbe et al., 2002). The phyto-geographical region, wherein the rainforest fragments of Benin are located, is the Guineo-Congolese zone (Adomou et al., 2011; Chatelain et al., 2004; Hawthorne & Jongkind, 2006; UICN, 1996) with two widely separated rainy seasons. The study site Drabo Gbo (6°30′N; 2°18′E; 60 m asl) has a mean annual rainfall of around 1200 mm with maxima in June and October. Mean temperatures are around 28°C, with maxima of 38°C in March-April and minima of 20°C, rarely down to 16°C in January.

The Village

Drabo Gbo is a densely built village of about 500 inhabitants, 30 km north of Cotonou, 12 km from the spreading town of Calavi and, in 2023, still 6 km away from the nearest paved road (Figure 1). Today, most of the surrounding land has been sold by the villagers to Benin citizens (another 1000 persons) flocking to the south. Many constructed small houses, using the rest of the land for farming and animal husbandry.

Figure 1.

Situation maps: Africa, Benin, southern Benin, study area in 2003 and 2023.

The Inhabitants

Detailed information about the inhabitants of Drabo Gbo and neighboring villages has been gained by the first author, who lives in the sanctuary for the last 28 years, through discussions within the vodun group of 10-50 men meeting weekly at the Legba square, participation at private invitations and funerals, and all the friends who help preserve the forest and its monkeys (described in details in fictionalized form in Neuenschwander, 2020).

Most inhabitants of the old village of Drabo Gbo farm on land they already sold, rear animals at home, and exercise various professions. For additional support, many work in town as motor taxi drivers. Many houses have wells to drag water manually from the water table at a depth of 25 m. Families often have ten children. Electricity is supplied through flimsy private lines connected to the country’s electric grid 1 km away in Drabo Kpevi. The official primary school is also in Drabo Kpevi; but many villagers send their kids to private schools in and around Drabo Gbo. According to observations by the first author, most if not all inhabitants live below the UNDP poverty line of 1-2 $ per person per day, despite economic improvements in the country (OECD, 2010).

In 1995, most villagers adhered to three different vodun cults, whereby most males – including the first author – were initiated to all three of them: 1- Zan gbetos, i.e., the hunters of the night, practically the village police, who perform their meetings and rites at the central square at the statue of a divinity called Legba; 2- the Oro, regionally active, with their main center the forest called Orojamè; 3- the revenants or Egungun, a cult from Nigeria that was introduced into the village in the 2010s. Since then, many young and a good part of the new settlers have been attracted to several evangelical Christian sects. In addition, a Muslim Mosque has recently been constructed, financed by Golf States, though villagers of this faith are few.

Creation of the Monkey Sanctuary

The monkey sanctuary was founded in 1995, when the first author bought 2.5 ha. of teak plantation and agricultural land from the elders of Drabo Gbo. At this time, almost no houses were found outside the village. Subsequent purchases over the next eight years are detailed in Figure 2 and Neuenschwander and Adomou (2017).

Figure 2.

Aerial view of Drabo Gbo, from Google Earth Engine (top). Map of the ‘Sanctuaire des Singes’ at Drabo Gbo (bottom), with GPS data for entrance doors, year of purchase and start of forest management and major clearings1 nursery-garden 1997 (house constructed 1997-1998); 2 papa-garage 1999, 2000; 3 Lissanou 1999–2003; 4 mill 2000; 6 Cooun 2001, part of Cooun cleared in 2010; 7 corridor-Dansou 2004, 2010; 8 Emile 2001–2007 partly cleared 2012; 9 ‘Maison de Jeunesse’ (MdJ house constructed 2005) 1998, cleared in 2013; 10 Tofinou 1998-2000; 11 Pierre 1999–2001; 12 Kakpo 2004; 13 Grande Forêt 1996, local fire in 2012; 14 AgoXwè 2000–2003; 15 Corridor north 1998, 2002–2003; Orojamè 1998; Fanto 1998–2000, partly cleared 2014; Dodja 2011, partly cleared 2016. Natural forests grey with border line, wood lots light grey with border line, compact villages light grey, unsurfaced roads as lines (see also Neuenschwander & Adomou, 2017).

The sanctuary comprises 14 ha (Figure 2). This includes a big block of land adjacent to the village, 2.5 ha of forest away from Drabo Gbo, plus two sacred forests (Figure 3) with old trees, namely the so-called Orojamè, the sacred forest of the Oro sect, and the sacred forest of Dodja. The Orojamè was bought by the first author on request by the Oro adherents in Fanto, who had lost control of their forest, because the other half of their family, recently converted to Christianism, had been given legal right over this land, and subsequently decided to cut all trees. The Orojamè and similarly the sacred forest of Dodja were handed over to the relevant vodun sects for ceremonial use.

Figure 3.

Sacred Forests of the ‘Sanctuaire des singes’, tiny islands of biodiversity: Orojamè (top), sacred forest of Dodja (bottom).

In 2014, all title deeds were given to IITA, which integrated the sanctuary in its government approved research program concerning biological control and the protection of biodiversity (Neuenschwander et al., 2023).

Data Source and Analysis of Maps

From satellite pictures, maps of increasingly greater accuracy were available for 1986, 2003, 2015, and 2023. The data used for maps (Figures 1 and 6) are Shapefiles obtained from the IGN Benin topographic base and SPOT images of 2023. The GIS tool used is QGIS Remote Sensing. Land cover maps were produced by visual interpretation, using false colours for grouping pixels into different land cover classes (Ahononga et al., 2021; Sikuzani et al., 2019). Processing SPOT images of Abomey-Calavi was based on assigning homogeneous zones to land use classes according to their spectral signature (Hammi et al., 2007). Visual interpretation was facilitated by Google Earth archive images of 2003 and randomly surveyed field reconnaissance points with geographical coordinates for 2023.

Management of the Forest

The forest is protected by a 3 m high wall along the village. The rest of the boundary is marked by five lines of barbed wire, i.e., easily accessible for animals as well as people who insist to penetrate illegally. Official access to the forest is only through the staff managing the sanctuary. They include two half-time guards and two villagers with minor tasks. Hunting is forbidden and collecting medicinal herbs, oil palm regimes, fire and construction wood, including bamboo, is only allowed under control by the guards. Trespassers are first reminded by the village elders, if necessary, punished by IITA’s security, or even transferred to the local foresters and, in a last step, handed over to police and a court of law.

Management of the forest consists in freeing trees and saplings alike from climbers, a task that is necessary almost weekly. The indigenous oil palm (Elaeis guineensis) has sometimes become locally dominant, as it has been protected through previous slash and burn cycles. Since such abundance menaces the biodiversity, several times up to fifty trees have been felled and sold for palm wine production and eventually the distillation of the local gin, Sodabi. Such clearings and those offered by trees toppled by storms offer sunlit habitats for planting out seedlings from the nursery.

Growth and Survival of Trees

In order to evaluate rapid growth in a fallow of originally 1 year (Figure 2) across the entire forest at Drabo, 404 well established trees (20 cm to several m high) of 63 species were chosen every 5m along the established paths and tagged in 1999. Height (in m) was measured with a tape and, on larger trees, estimated by comparison with a person standing at the base of the tree. From 2004 onward, the circumference of the stem at breast height (in m with two decimals) was also measured with a tape (see Table 4 Supplemental Files).

In this data set, many species are represented by only a few trees. Many trees disappeared during this period or could not be found again. For the evaluation, therefore, only the fast-growing trees of three genera are retained, namely 25 Albizia of 3 species, 14 Antiaris, and 36 Blighia of two species. Some trees were reintroduced into the counts in the course of the years; but are eliminated from the evaluation. Mortality of trees over the entire period was computed as mortality per year. Trees broken, presumably by human intervention, were counted separately.

Introducing Plants from Other Forests and Protection of Rare Trees

The introduction, survival, and loss of plants was continuously monitored and documented (Table 5, Supplemental files) by the first author during 6-8 months per year from 1996 to 2024, with identifications by AA and HD. We describe and quantify the condition of different functional groups of plant species, compare their change in time, and evaluate the survival of newly-introduced, threatened plants. In addition, the species in the two sacred forests, those of Dodja and the Orojamè, and in the triangle of Drabo Fanto, are compared with species that grew spontaneously in the teak forest and fallow of Drabo Gbo, where the influence of the local population was strongest.

Numbers of species with different growth patterns or survival are expressed in percentages of the total number of species. Comparisons of loss and decline are evaluated by Chi-square tests at p= 0.05 (marked with *).

During excursions into the remaining rainforest patches in southern Benin (localities listed in Table 5, Supplemental Files and below), seeds and seedling were collected in plastic bags and transplanted to large pots in the nursery or directly into the forest to locations that were deemed to correspond to the collection site. In the last two years, the roots of uprooted plants (seedlings and saplings) were immediately wrapped into moist towels to improve survival during transport.

A list of all species registered in these forests by July 2023 is given in Table 5 Supplemental Files. Species from aquatic, semi-aquatic, coastal-sand, or rocky habitats as well as horticultural species are excluded. The Supplemental File lists 659 species, of which 589 are alive. One more species, Oxyanthus tenuis, was detected in Dodja in January 2024, but not included in the evaluation. The file has the following headings:

– Taxa (Pteridophytes as 1Pteri (only 1 species alive), Gymnosperms as 2Gym (none survived), Monocotyledons as 3Mono (a total of 69 living), Dicotyledons as 4Dicot (a total of 519 living), and non-identified as 5nonid (all are Dicotyledons).

– Family names followed taxonomic revision APG 1 (1998) in previous studies (Adomou et al., 2011; Neuenschwander & Adomou, 2017) in order to have the least differences with Akoègninou et al. (2006). Here, we use the most recent revision (APG IV (2016). Sub-species may be named; but are counted only as one species (e.g., Mansonia). Species that could not be identified but had a clear taxonomic, growth and survival status were marked by ‘?’ and included in the tabulations. Where different, the names used in Neuenschwander and Adomou (2017) were added in brackets ‘[…]’.

– Species that are either new or not included in Akoègninou et al. (2006) are marked as a. A total of 12 species falls into this category, two of them since lost.

- Species names followed by x are registered in inaturalist under PeterNeuenschwander.

– Plants are described as parasites/saprophytes (1 species), epiphytes (4 spp.), herbs (150 spp.), climbers (155 spp.), shrubs (64 spp.) or trees (215 spp.) based on the description in Akoègninou et al. (2006).

– The origin of the species, named chorology, is indicated according to Adomou et al. (2006; 2010; 2011) and Akoègninou et al. (2006) as follows (map: Figure 2 in Neuenschwander & Adomou, 2017) and grouped as follows: Group I includes species with large distributions: At Afrotropical species with distributions beyond West and Central Africa into Madagascar, PAL Paleotropical, and Pt Pan-tropical species, i.e., all species that have penetrated or invaded West Africa from other floristic regions, for example coconut, teak, etc. West African species that have similarly spread across the world are indicated by their original zone, for example oil palm. Group II encompasses species from areas with only one rainy season: SG Guineo-Sudanian transition zone species, SZ Sudanian savannah species, and S Sahel savannah species. Group III encompasses species that are adapted to the local climate with high rainfall and two rainy seasons: GC Guineo-Congolian forest species that are distributed across the Upper and Lower Guinean and into the Congolese zone east to Sudan, Uganda, Kenya, GO Upper Guinea forest species from west of the Dahomey Gap with an eastern limit in Benin or nearby Nigeria, GE Lower Guinea forest species from east of the Dahomey Gap with a western limit in Benin.

– The next columns indicate where the transplanted species had been collected, such as Ahozon, the only remaining coastal forest in the Dahomey Gap (about 10 km south of Drabo), Ouéga and the IITA campus a few km south of Drabo, Agongbè north of Drabo, Hévié north of Pahou, Avrankou on the Iguidi River near the Nigerian border, Dangbo north of Porto-Novo, and Tanougou waterfalls outside the Penjari Park in the Sudan savannah, together with their location, where they were transplanted in the study forests (Figure 2), the years of collecting and transplanting, and-separately-in which form they had been collected, namely seeds, small plants, or sticks.

– Abundance in 2016 was ranked as follows: 1 = 1–4 plants established; 2 = 5–10 plants; 3 = up to 20 plants; 4 = common species; 5 = abundant species. Here, the definition was enlarged for herbs: 1 = occasional occurrence in openings, still common outside forest.

– The maximum height and circumference were assessed in January-February 2023 in all study forests as described above. For most species, biggest trees were found in Drabo; in addition, especially big specimens from Dodja (Do), Fanto (Fa) or Orojamè (Or) are indicated.

– The population trend is roughly estimated as s = stable, i = increasing, d = decreasing, c = cut, or L = lost by July 2023. d thereby means that the species became less numerous and/or individual plants became sick or stressed. Plants planted in May-June 2023 were assigned the status of (s), or assumed to be doomed (L). Species collected once only as seeds that did not sprout, or sticks that did not root, were excluded from the list (contrary to Neuenschwander & Adomou, 2017).

– Red List status was given according to Neuenschwander et al. (2011) and IUCN (2016) based on IUCN criteria as NT = near threatened, VU = vulnerable, EN = endangered, CR = critically endangered, EW = extinct in the wild in Benin. All other species were considered as LC = Least Concern, i.e., not threatened or not assessed.

– Suspected reasons for difficulty in establishment are given as: - dormant seeds, which were sometimes treated with hot water or by scarifying to break dormancy, - drought, i.e., temporarily too little water, - savannah species, of which we suspect that they do not support transfer to two rainy seasons and are therefore not capable of reaching the coast, - medicinal use; often the roots of these species are harvested for increasing male potency.

– The number of samples, which include one to maximum 10 plantlets or seeds per species and date, is indicated. Repeated replanting of sticks (e.g., Rhodognaphalon) from trees collected in Drabo are not considered new introductions and are excluded from these counts (contrary to Neuenschwander & Adomou, 2017). The status of 2023 is compared with the one of 2016, when the plant list was last assessed (Neuenschwander & Adomou, 2017).

– Plants originally found on the 14 ha and in their vicinity of one hundred meters are marked with x. Some of them were also reproduced and transplanted to other sites.

Impact of Climate Change

In order to assess possible impact of climate change, weather data from Drabo and the IITA station were compiled over 25 years (Figure 7). Monthly mean rainfall (mm) and temperature (^oC, mean, mean maximum and mean minimum) were recorded daily in an official weather station at the IITA research station. In Drabo Gbo, only rain was recorded (with a rain gauge).

Apart from higher temperatures and changes in rainfall, climate change also leads to more severe extreme climate events (IPCC, 2022). The impact of thunderstorms was quantified by the number of big trees (>20 cm diameter) that were either broken or toppled (with the root disk turned up). Naturally dead trees, particularly the many oil palm trees that die each year, were excluded. So were the small trees and shrubs that were damaged when the big trees crashed onto the forest floor. Data are limited to the 4 ha of Cooun (Figure 2), which received the bulk of introduced plants, and papa (Table 7, Supplemental files), because this is an area of common history starting out as meadow or young fallow.

Conservation in Action

The livelihood of villagers next to protected land is important for its conservation (Neuenschwander & Sinsin, 2011; https://www.iucn.org; https://www.birdlife.org). The first author, with support by IITA and the Leventis Foundation, assisted villagers by co-financing the construction of 15 toilets to avoid spoilage and damage in the adjacent forests, and the construction of wells. A ‘Maison de Jeunesse’ was built as a community centre. For several years, a mill was maintained; but later abandoned when transportable private mills were introduced. Children were offered additional schooling combined with eco-tours in the forest. Four MSc and one PhD student with nature conservation topics were sponsored at local universities. Privately, microcredits were offered also to women. Under supervision by the guards, low level exploitation of the forest for fire wood, timber, and non-timber products like medicinal plants and snails was allowed.

Promoting Ecotourism

The protected forests are clearly visible on Google Maps under ‘Sanctuaire des singes’ (including GPS data) (Figure 2). Visitors, guided by these data, are led through the forest by PN or guards and shown red-bellied monkeys and forest vegetation. For a 2-hour visit, foreigners pay a modest entry fee of about $6, Benin citizens $3, half for children, in order to support maintenance. For inhabitants of local villages, visits are encouraged and free. Information is available on Facebook managed by IITA staff and assistant teachers: https://www.facebook.com/people/Sanctuaire-des-singes-de-Drabo-Gbo-de-lIITA-B%C3%A9nin/100081911083232/

Results

Regrowth and Survival of Common Trees

A first overview of the changes in the vegetation from early fallow/crop fields to a forest with closed canopy is given by Figure 4.

Figure 4.

Fallow field viewed from the edge of what became Grande Forêt eastward towards the big Cola gigantea tree in 1998 (top, by PN) and from 100 m above ground over the closed forest in 2024 (bottom, by Georg Goergen, IITA).

The speed of regrowth of saplings in the early phase of forest development after clearing is rapid and differs between species (Figure 5). It is most rapid for the light demanding and fast-growing Albizia spp. pioneer trees, which are the dominant fallow species.

Figure 5.

Growth of three common trees: top: Albizia (original N=25), middle: Antiaris (original N=14), and bottom: Blighia (original N=36). Diameter of points indicates circumference.

By 2023, the same species in the same locations reached maximum height and circumferences of 29 m and 2.20 m, respectively, for Albizia adianthifolia, 34 m/2.24 m for Antiaris toxicaria, and 38 m/3.10 m for Blighia unijugata.

Mortalities were assessed as follows:

- Blighia: Total mortality: 6 out of 36 trees = 16.7% in 13.1 yrs = 1.3%/yr, natural mortality: 3/36 = 8.3% in 13.1 yrs = 0.6%/yr;

- Antiaris: Mortality: 5/14 = 35.7% in 11.9 yrs = 3.0%/yr, natural mortality: 2/14 = 14.3% in 11.9 yrs = 1.2%/yr;

- Albizia: Mortality: 4/25 = 16.0% in 12.6 yrs = 1.3%/yr; natural mortality: 2/25 = 8.0% in 12.6 yrs = 0.6% per yr.

In conclusion, yearly mortality was in the order of 1% per year.

Among all forest plants, the following make up the bulk of the forest (abundance of 5 in Table 5 Supplemental Files). All are local species:

- trees: Elaeis guineensis, Monodora tenuifolia, Holarrhena floribunda, Newbouldia laevis, Albizia adianthifolia, Albizia glaberrima, Albizia zygia, Millettia thonningii, Cola gigantea, Sterculia tragacantha, Antiaris toxicaria, Ficus exasperata, Blighia sapida, Blighia unijugata, Lecaniodiscus cupanioides, Pouteria alnifolia;

- shrubs: Agelaea pentagyna, Dichapetalum madagascariense, Mallotus oppositifolius, Carpolobia lutea, Chassalia kolly; and

- climbers: Artabortrys velutinus, Uvaria chamae, Cnestis ferruginea, Rourea coccinea, Reissantia indica, Triclisia subcordata, and Cremaspora triflora.

Trees are most species-rich with 215 spp. (= 36.5 % of all plants). They form a more or less closed forest canopy, with the following maximum heights in m (only trees above or equal to 20 m height are listed, * for introduced spp.): Cocos nucifera (height 20 m, circumference 1.08 m), Spondias mombin (24, 1.08), Holarrhena floribunda (20, 0.96), Celtis mildbraedii (>30, 1.60), Celtis zenkeri* (30, 7.10), Terminalia leiocarpa (20, 0.86), Terminalia mantaly* (28, 1.16), Terminalia superba* (22, 0.95), Tectona grandis (20, 1.24), Senna siamea (39, 1.24), Cassia sieberiana (20, 0.58), Erythrophleum suaveolens* (30, 1.64), Acacia auriculiformis (30, 1.13), Ceiba pentandra (36, 2.61), Cola gigantea (38, 12.00), Albizia adianthifolia (29, 2.20), Albizia ferruginea (27, 2.36), Albizia glaberrima (23, 0.61), Albizia zygia (20, 1.70), Leucaena leucocephala 30, ?), Samanea saman* (24, 1.35), Rhodognaphalon brevicuspe (35, 2.15), Triplochiton scleroxylon* (35, 1.26), Khaya senegalensis* (30, 1.26), Antiaris toxicaria (38, 2.23), Psydrax parviflorus (24, 1.44), Blighia unijugata (38, 3.10).

Among these 27 tree species with large trees, seven have been introduced. Overall, four of these species have abundance scores of 1; 4 have abundance scores of 2, i.e., up to 10 plants; but the remaining 19 species all have abundance scores of 3, 4, or 5.

Among all plant species, 56.6% (331 out of 589) are represented only by 1 to 4 specimens (abundance score 1).

Loss of Species

The loss of species from the ‘Sanctuaire’, as extracted from Table 5 Supplemental files, is indicated in Table 1.

Table 1.

Loss of plants of different origins.

Origin		lost	Total	%lost per line
I	Pt	15	151	9.9
	PAL	1	18	5.6
	At	5	80	6.3
II	S	6	14	42.9
	SZ	7	25	28.0
	SG	10	93	10.8
III	GC	19	244	7.8
	GO	0	16	0.0
	GE	1	7	14.3
Non-id		0	5	0.0
Total		64	653	100.0

Overall, the loss of species is rather modest (64 out of 653 = 9.8%). There is a marked loss of plants from drier areas with only one rainy season (Group II: S, SZ, SG 17.4% vs. Group I: Pt, PAL, At 8.4%: Chi square 6.83*, and vs. Group III: GC, GO, GE 7.0%: Chi square 10.73*).

The loss of the 64 species according to growth form, is as follows: trees 41 (16.1% of all trees), shrubs 4 (5.9%), lianas 9 (1.4%), herbs 8 (5.1%), others 2.

As is to be expected, species that disappeared, had only small populations. Among the 64 lost species, all had abundance score 1, i.e., had only maximum 4 plants, or had been introduced many times without success. Locally occurring plants were rarely lost (9 out of 64 = 14.1%).

Most losses occurred in the year the species was first planted out into the forest. The following species were, however, collected and planted out in the forest on 5-6 occasions over many years and still died out during the first year: Platycerium elephantotis, Myrianthus arboreus, Pterocarpus erinaceus, or Rhigiocarya racemifera. There were, however, notable exceptions like Prosopis africana (SZ), Lophira lancoelata (SZ), Parkia bicolor (GC), which were seemingly established, some even several times and for several years, but eventually died out in the sanctuary.

The main reasons for losses are: unfit material when collected and transported, as well as scarcity of rain or humidity or excessive humidity for savannah species (Pterocarpus erinaceus, Prosopis africana, Lophira lancoelata).

New introductions were made every year. Since 2016, a total of 220 samples was introduced from outside the sanctuary. In 161 cases (= 73.2%) at least one plant of the same sample survived. These introductions involved 159 species, among which 21 (= 13.2%) did not survive. For the further analysis, lost plants are excluded.

Freeing trees from lianas was routinely performed. Otherwise, small trees were smothered by lianas, as was often observed wherever a tree had been broken or fallen, whereupon the opening became covered with Combretum spp., Dioscorea spp., Mezoneuron benthamianus, Reissantia indica, and rarely others.

During storms, Albizia trees covered with arm-thick Cryptolepis nigrescens were often felled (see below).

Inventory of the Flora

Recent introductions include 24 species that have not been recorded before. On balance, since 2016, 20 spp. have been lost and 24 new spp. have been added, bringing the total to 589 live species among which 257 are introduced.

56.4% (332 out of 589) of all species were already present in Drabo Gbo, but many of them were still multiplied and transferred to other parts of the forest. This transfer involved 44 samples of 20 locally established species, among which cuttings from only 2 samples did not survive.

Sometimes, establishing new plants was not easy. After many failed attempts at establishment, we finally succeeded in obtaining viable plants of Maesopsis eminii (after 5 attempts), Macaranga barteri (after 8 attempts over 9 years) or Pierreodendron kerstingii (after 4 attempts); all these species exhibit strong affinity to humid soil. Establishment succeeded only because young plants were irrigated by hand, sometimes daily, during the second half of their first dry season.

Most introduced plants remained at a population size of 1-4 individuals (abundance score 1). There were, however, plants that became reasonably abundant (abundance score 2 or even 3) following sometimes many introductions: Terminalia superba (yearly introductions), Rhodognaphalon brevicuspe (yearly introductions), Maranthes robusta (12 introductions), Baissea axillaris (8), Spondias mombin (5), Barteria nigritana (8), Leptactina arborescens (5), Psydrax parviflora (5), Acridocarpus alternifolius (7).

Many species (33.8%, i.e., 199 out of 589 spp.) declined in abundance and vigor. Shading by the growing forest was the reason, most notably for Trema orientalis and Chromolaena odorata. Herbs were particularly affected; 69.3% disappeared from the central part of the forest and survived only in openings and at the edge, as detailed in Table 2.

Table 2.

Population dynamics for species of different growth forms. N = number of species.

Population	Decline	Stable	Increase	Total	% Decline per line
Trees	55	148	12	215	25.6
Shrubs	14	47	3	64	20.6
Climbers	25	119	11	155	16.1
Herbs	104	39	7	150	69.3
Epiphytes	0	4	0	4	0.0
Sapro/para	1	0	0	1	50.0
Total	199	357	33	589

Plants of different origins (chorology) (Table 3) were represented as follows:

Table 3.

Chorology vs. survival.

	Decline	Stable	Increase	Total	%decline of total line
Group 3: Pt, PAL, At	141	78	9	228	61.8
Group 2: S, SZ, SG	40	63	6	109	36.7
Group 1: GC, GO, GE	18	211	18	247	7.3
Non-identified	0	5	0	5	0
Total	199	357	33	589	33.8

Adapted local species (Group III: GC, GO, GE) accounted for 41.9% (247 out of 589), those from nearby drier areas but with only one rainy season (Group II: S, SZ, SG) for 18.9% (109 out of 589), and the wide-range species from far away origins (Group I: Pt, PAL, At) for 38.7% (228 out of 589) of all species.

Their losses differed a lot. The exotic generalists showed highest declines (141 out of 228 = 61.8%); species from drier areas (S, SZ, SG) showed less decline (36.7%: Chi square 6.03*), while rainforest species had far smaller numbers of species in decline (7.3%, Chi square comparison to S, SZ, SG: 6.03*).

As can be expected, species that are only represented by 1-4 specimens suffered higher rates of decline. Among the 199 species in decline, 89.9% (179) had an abundance score 1. By contrast, among the 390 species with stable or increasing populations only 39.9% (152) had abundance score 1 (Chi square 139.09*).

Protection of Rare Trees

Among the 589 plant species living in the ‘Sanctuaire’, 58 are rare and threatened (Table 6, Supplemental). With the exception of 9 species, all have been introduced. 21 species are represented only by 1-4 specimens (abundance score 1); but many have become reasonably common and of good size. The locally present Khaya senegalensis and Milicia excelsa reach 30 m or more; among the newly planted trees, Celtis mildbraedii and one Triplochiton scleroxylon also reach 30 m.

Twelve species are critically endangered and with one exception, all have many specimens (abundance score 2 or above) and attain considerable sizes. Of particular interest is Diospyros barteri, discovered inside the sacred forest of Dodja.

Medicinal plants like Mondia whytei, Acridocarpus alternifolius, and Turraea heterophylla are well established and are not rare.

Several other uncommon species are noteworthy: Englerophytum oblancoelatum, Ficus mucuso, Maesopsis eminii, Napoleonaea imperialis. Oxyanthus tenuis was only identified in Dodja in 2024, which brings the total list of living plants to 590.

Comparison Between Old Sacred Forests and the Newly-Constituted Forests

Each separate forest of the sanctuary harbours unique species of trees and shrubs (herbs and lianas are not included as they are often too transient and difficult to find). The sacred forest of Dodja has Celtis mildbraedi, Connarus africanus, Diospyros barteri, Oxyanthus tenuis, Stereospermum kunthianum (just outside), Salacia longipes, and Zanthoxyllum leprieuri, which are all originally lacking in the other forests. The Orojamè has Hildegardia barteri, Psydrax parviflorus, Triplochiton scleoxylon (just outside), Vepris verdoorniana. The triangle of Fanto has Crossopteryx febrifuga, Gardenia ternifolia, Ochna schweinfurthiana, Ximenia americana. Drabo Gbo has 20 tree species that were planted by the resident population as fruit and fire wood trees or for medicinal purposes. In addition, Diospyros abyssinica, Ficus mucuso, Loeseneriella africana Premna angolensis, Salacia pallescens, Trichilia megalantha, not recovered in other sites, are original forest plants. Most of these species were later transplanted also to other sites within the sanctuary.

Influence of Common Drivers: Urbanization

In 1986, Drabo Gbo consisted of a densely populated village surrounded by agricultural and fallow land. 65.0% of the land was covered by shrub savannah. By 2003, fields and fallow fields with or without palm trees became predominant; shrub savannah had diminished to 0.8%. The two large units of open forest and tree savannah identified in 1986 all disappeared by 2003. The area covered by fields and palm trees increased from 19.9% to 84.1% from 2003 to 2023 (Figure 1).

In 1995, when the Sanctuaire was created, the only rainforest vegetation was found in the sacred forests of Orojamè and Dodja, plus near the big Cola tree at the Legba square. 28 years later, natural forests in the form of the ‘Sanctuaire’ have become a species-rich secondary forest and are clearly visible even on satellite pictures (Figure 2).

Urbanization in the form of small houses on small agricultural lots increased particularly in the last few years (Figure 6). Between 2015 and 2023, the number of houses outside the village grew from 3607 to 5108 units. The residential area increased from 15.7 ha in 1986 to 282.4 ha in 2023. This is the result of the immigration of people looking for land for housing near the urban centers of Abomey-Calavi and Cotonou, where they work.

Figure 6.

Building dynamics in the research area: 2003 left and 2023 right.

This development is still vivid in the memory of old villagers. It was learnt that, 80 years ago, what is now called Cooun (Figure 2), was a dense forest feared by children. About 60 years ago, Grande Forêt was cut and replanted with teak. Another old villager recounted how, in his youth, he had hunted red-belied guenons in these forests.

Increased urbanization means that more people need fire and construction wood, medicinal plants, and food (snails, game). People collecting snails, ground squirrels or firewood were often stopped and reminded that collection was allowed only under supervision by a guard. Fortunately, none of the, by 2023, over 40 red-bellied guenons, nor any of the rather common duikers has ever been shot.

Each year, several, mostly common trees are vandalized. The thieves are punished by vodun elders or IITA security. Only one case of attempted theft of land has been prosecuted by the court. Since 2013, damage to maize crops and fruit trees by monkeys is another yearly complaint. These losses are compensated for. In addition, neighbours complain about overhanging branches, which are then cut to protect nearby houses.

Visits by children in classes offered by assistant teachers in the Maison de Jeunesse twice a week are well received. Visits by local schools are the exception, while international schools from Cotonou visit every year. Their reports on Google Earth and on Facebook give high approval ratings of 4-5. Thus, in 2023, on 23 occasions, 180 visitors were given a guided tour and paid a total of CFA 235 000, i.e., about $350 or 4 months of the official minimum salary SMIG.

The local vodun cults appreciate the protection given by the ‘Sanctuaire des singes’ to their holy forests, where yearly celebrations are held. Due to the death of the regional vodun priest, the Vodunno, and ill health of the Zan-gan, the chief of the Drabo Zan-gbeto cult, vodun celebrations have markedly declined in the last three years. Moreover, the increasing influence of evangelical Christians, who dislike and counteract the vodun elders, makes protection of the sanctuary ever more challenging.

Influence of Common Drivers: Climate Change

Over the last 25 years, the total annual rainfall stayed rather stable, but with important variations from one year to another. Despite the short distance of about 10 km between the two stations at IITA and in Drabo Gbo, differences in rainfall were often substantial, though without any clear trend (Figure 7).

Figure 7.

Total annual rainfall in mm for Drabo Gbo (6^o30’N; 2^o18’E; solid line) and the IITA station Calavi (06^o26’15’’N; 02^o19’42’’E; broken line) from 1999 to 2023.

In the rather uniform forest of Cooun and papa (Figure 2), a total of 53 big trees were affected by storms in the last 6 years (Table 7, Supplemental files). Before, no storm damage had been observed, except for the toppling of a sacred Cola gigantea tree and the breakage of a big branch of the big Cola tree (Figure 8), both before the year 2000. From 2019 to 2023, 15 trees were affected, 40% toppled and 60% broken. With the exception of 3 trees, all were Albizia. This amounts to 0.75 trees per ha per year.

Figure 8.

Big Cola gigantea tree (left) and storm damage 2024 (right).

During three thunderstorms in the first half of 2024, another 38 trees were hurt, about half broken, half toppled (Figure 8). All were Albizia, with the exception of 5 trees, which were affected by a nearby toppled Albizia tree (14 were A. adianthifolia, 19 A. zygia). This amounts to 9 trees per ha per year, a twelve-fold increase over the previous 5-year period. The same storm also toppled a 27 m high A. ferruginea tree with a circumference of 2.36 m (measured in 2023) on Emile; but left the 38 m high Cola gigantea with a circumference of 12 m undamaged (Figure 8). All toppled Albizia trees had astonishingly small root discs of 3-5 m diameter with numerous finger-thick roots and only a few roots of 5-8 cm diameter.

Discussion

The tiny forest fragments in Benin (Adomou et al., 2011) cannot each harbour all the rich Congolian and Upper Guinean flora. 14 threatened species are only known from one or two sites (Neuenschwander & Adomou, 2017). Similarly, many species - before being distributed further - were only found in one of the four forest sites of the sanctuary. Even within the same forest, species like Oxyanthus racemosus and Pavetta corymbosa, now common across the sanctuary, were represented by only one plant each in the first evaluation in 1996. The claim of reconstituting a rainforest therefore needs clarification. The sanctuary in Drabo Gbo does not represent any original condition, but has more species than any other forest in southern Benin, reuniting species that do not grow side by side anywhere else in the country, except in a botanical garden.

A comparison of the list of species determined in 2016 (Neuenschwander & Adomou, 2017) and the one of this text in 2023 is complicated by the fact that 51 species were assigned a different name according to the newest nomenclature (APG IV, 2016). For the same reasons, 65 species, not necessarily the same, were attributed to a different family.

The same comparison reveals a loss of 20 species and a gain of 24 species during the last 7 years because of new collections, with a new total of 257 introduced species. It is interesting to note that despite this large number of introductions, no species ever showed any sign of being invasive. Exotic species like Leucaena leucocephala, Senna siamea and Acacia auriculiformis, which have originally been introduced as alley cropping trees (Mutsaers et al., 2017), could not compete in the growing canopy and were relegated to the edge of the forest. Overall, all introduced species account, however, for only 1.7% of all plants growing in the sanctuary (Neuenschwander & Adomou, 2017).

A total of 64 species were lost after having been transplanted to the forest. Most were from drier areas and all were implanted only in low numbers. One third of all species declined in abundance or vigor. This is mostly attributable to increased shading due to the natural succession from rather open fallow land to a forest with a closed canopy. Most reductions concerned herbs, which often survived only in new openings due to tree fall, at the very edge of the forest in adjacent private compounds. Similarly, Trema orientalis reappeared from soil-stored seeds wherever a large opening appeared, but died after some years when its trees were shaded out.

Chromolaena odorata, which previously formed dense thickets, disappeared mostly from the forest. This reduction is partially attributed to the spread of two biological control agents, that had been released in Ghana in the 1990s, namely the arctiid moth Pareuchaetes pseudoinsulata Rego Barros (Timbilla & Braimah, 2000), and the tephritid fly Cecidochares connexa Macquart in the 2000s in Côte d’Ivoire (see also Chikoye et al., 2011; Day et al., 2013). Both agents were observed in the sanctuary and specimens deposited in the IITA Biodiversity Collection.

It must be expected that, due to successional changes, alpha-biodiversity, expressed as the total number of species, will decline as seen in other systems (Barlow et al., 2007). In addition, stochastic loss of species because of the small size of the sacred forest refuges and here of the ‘Sanctuary’ (Fahrig, 2003; Klein et al., 2014) remain a threat.

Several threatened plants, despite initial success, could not establish; others were established only after years of trial and error. Better planting material certainly played a role, but unfortunately, in some cases, we do not know why plants could eventually establish and grow.

By 2023, the sanctuary harbored 58 rare and threatened species that, with 3 exceptions, are on the country’s Red List – more than in any other forest in Benin (Adomou et al., 2011; Neuenschwander & Adomou, 2017). Some have grown to big trees. Twelve species are considered critically endangered, mostly well represented by several specimens. Many threatened species that have now found a sanctuary in Drabo are known only from one or two other locations in Benin (Neuenschwander & Adomou, 2017). Several other rare species that are established in the sanctuary, still need to be included in the IUCN Red List. In addition, two climbers that are established since several years and one local tree could not yet be identified even by botanists from the ‘Musée royal de l'Afrique centrale’, Tervuren (Belgium). They are probably new species for Benin and all of them eventually merit a place on the IUCN Red List for Benin.

Of particular interest is Diospyros barteri, a tree that has been newly discovered for Benin in the sacred forest of Dodja, which is part of the sanctuary (Dassou et al., 2024). This find indicates how little the flora of these small sacred forests, protected by vodun cults and of difficult access, are known. Because of its apparent rarity, we catalog this species as critically endangered at national level, while it is categorized VU by the international IUCN list (https://www.iucnredlist.org/en).

Some species were only established after numerous trials, reminding one of the saying that rare species are not rare for nothing. In many cases, treatment of seeds and the necessary care of roots and their accompanying fungus species still needs a lot of research. Further collections of up to now failed introductions are therefore warranted.

While new species were introduced, rare species from the nursery were also offered to the botanical garden of UAC and to NGOs active in forest conservation, an exchange for the better survival of localized species, as recommended in the Red List (Adomou et al., 2011).

Forest maintenance, including pruning of common species to ensure access to light, is important to preserve rare trees. This included the continued removal of dense stands of oil palm trees, which blocked natural succession. Removal of lianas on a weekly basis is considered to be key for the success of plants that have been transplanted from the nursery. We would therefore caution all the NGOs that establish nurseries of easily reproducible indigenous trees in Benin, that the main labor in forest rehabilitation, according to our experience, is not the establishment of nurseries, but the care of plants once they have been transplanted into the forest.

By planting in clearings due to natural tree fall and mostly without irrigation, we employed traditional forestry practices used for rehabilitation (Chazdon, 2014; Sabogal, 2007; Stanturf et al., 2012). Thanks to deep soils and by adding species, the Drabo forests now reach the semblance of a secondary forest. As the only rehabilitated forest in Benin, the ‘Sanctuaire des singes’ can serve as a model for better management of many of the threatened sacred groves in the Ouémé Valley, a UNESCO World Heritage site.

The sacred forests in Benin indeed are small; but deforestation, through forest fragmentation due to tree felling and acerbated by climate change (Corlett, 2014), is pervasive even in the Upper Guinea forests of Côte d’Ivoire and Ghana. There, forests of 4 ha or less now represent 64% the forest cover (Chatelain et al., 2004), i.e., not so much different from Benin. Similarly high deforestation rates are reported from Nigeria (Giresse, 2008).

In the sanctuary, arrests for felling trees and dragging them through the forest, thereby damaging newly established plants, increased. Policing by the vodun elders was not sufficient and IITA’s security force sometimes had to intervene. Offering free access to the forest for the collection of medicinal plants was rejected, because first finders would uproot the whole plant instead of sustainably harvesting the needed plant parts. Medicinal plants are still exploited, but only under supervision by guards.

Fortunately, climate change did not affect the sanctuary until 2023. A comparison of rainfall between IITA and Drabo Gbo, gives no clear trends. Though some rains became more erratic, the overall yearly rainfall remained rather constant. For a possible mitigation of climate change attributable to forests (Corlett, 2014), the forests of the ‘Sanctuaire des singes’ are evidently too small to show any effect to benefit the local population.

The Sixth Assessment Report of the Intergovernmental Panel on Climate Change (2022) also warns of increased thunderstorms with high winds as a consequence of a heating world. The twelve-fold increase in tree-toppling and tree-breaking observed in 2024, is therefore highly worrying. With the previous rate of <1 tree lost per ha per year, such damage could be naturally sustained; but with 9 trees per ha lost in 2024, the holes in the forest canopy become dangerously big. To maintain a healthy forest requires ever more replanting and suppression of lianas. By far most of the damage occurs on early-succession trees, mostly the four Albizia species. A speeded-up succession with Antiaris, Cola, Celtis, Chrysophyllum, Diospyros, Triplochiton, Zanthoxylum, etc., all of which did not suffer any damage, is therefore asked for.

Eco-services (Kuyah et al., 2016; Rowland et al., 2016), like forest foods contributing to healthy diets, are exploited at a low level, and local acceptance of the sanctuary is considered reasonable (Bello-Bravo, 2020). Nevertheless, in a village of 500 mostly poor people, intrusions and conflicts cannot be avoided. It is evident that this balancing act between nature conservation and support to villagers becomes ever more difficult under increasing urbanization. We must assume that the offered benefits like improvement in infrastructure, health and finances impacted the livelihood of many villagers and avoided possible damage to the sanctuary.

Protecting all wildlife, not just plants, remains a challenge (Mansourian et al., 2005). This forest provides a sanctuary also to five species of primates, four species of martens, one antelope, 80 bird species, 10 snake species, two chameleons, and numerous insects, which make up for a living forest (Neuenschwander et al., 2015). This 14-ha reserve represents two dozen sacred forests, which effectively protect biodiversity in a human-impacted landscape (Willis & Birks, 2006).

Contrary to sacred forests, the ‘Sanctuaire des singes’ is open to the public. Ecotourism is alive and most visitors are highly enticed by the forest walk and the contact with the red-bellied guenons (Neuenschwander, 2024). The economic data show, however, that ecotourism’s claims for a general improvement of village life are overrated for the situation in Drabo.

Like other African cities, Abomey-Calavi and Cotonou, Benin’s two largest towns, are spreading and start engulfing the ‘Sanctuaire’. This overpopulation and the attitude of many newcomers, who are looking for and demanding urban amenities, brings new challenges.

Implications for Conservation

Since the ‘Sanctuaire des singes’ harbours more threatened plants than any other natural forest in Benin - as well as a free-living population of critically endangered red-bellied guenon - it is of prime conservation value. Because it belongs to IITA and the first author supports local needs, the forest and its inhabitants survive – up to now. The biggest challenges consist in sustainably managing the creeping urbanization and increasing storm damage due to climate change.

Protection of the forest by IITA continues under the heading of biodiversity conservation for the benefit of a productive and sustainable agriculture (McNeely & Scherr, 2001; Neuenschwander et al., 2023). Similarly, over 150 field stations world-wide serve as earth observatories and biodiversity sanctuaries (Eppley et al., 2024) with high social benefits (Rasmussen et al., 2024). With its high educational and scientific value, the forest is recognized by universities and NGOs active in nature protection in the nearby Ouémé floodplain, a UNESCO World Heritage site, as a good example for managing sacred forests.

Supplemental Material

Supplemental Material - Managing a Reconstituted Rainforest Remnant and its Threatened Plants in Benin

Supplemental Material for Managing a Reconstituted Rainforest Remnant and its Threatened Plants in Benin by Peter Neuenschwander, Aristide Adomou, Hospice Dassou, and Clément Adjire in Tropical Conservation Science

Supplemental Material

Supplemental Material - Managing a Reconstituted Rainforest Remnant and its Threatened Plants in Benin

Supplemental Material for Managing a Reconstituted Rainforest Remnant and its Threatened Plants in Benin by Peter Neuenschwander, Aristide Adomou, Hospice Dassou, and Clément Adjire in Tropical Conservation Science

Supplemental Material

Supplemental Material - Managing a Reconstituted Rainforest Remnant and its Threatened Plants in Benin

Supplemental Material for Managing a Reconstituted Rainforest Remnant and its Threatened Plants in Benin by Peter Neuenschwander, Aristide Adomou, Hospice Dassou, and Clément Adjire in Tropical Conservation Science

Supplemental Material

Supplemental Material - Managing a Reconstituted Rainforest Remnant and its Threatened Plants in Benin

Supplemental Material for Managing a Reconstituted Rainforest Remnant and its Threatened Plants in Benin by Peter Neuenschwander, Aristide Adomou, Hospice Dassou, and Clément Adjire in Tropical Conservation Science