Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Journal Pre-proof
Ecological patterns and effectiveness of protected
areas in the preservation of Mimusops species’
habitats under climate change
Gisèle K.S. Sinasson, Charlie M. Shackleton,
Oscar Teka, Brice Sinsin
PII: S2351-9894(21)00077-9
DOI: https://doi.org/10.1016/j.gecco.2021.e01527
Reference: GECCO1527
To appear in: Global Ecology and Conservation
Received date: 23 July 2020
Revised date: 1 March 2021
Accepted date: 1 March 2021
Please cite this article as: Gisèle K.S. Sinasson, Charlie M. Shackleton, Oscar
Teka and Brice Sinsin, Ecological patterns and effectiveness of protected areas in
the preservation of Mimusops species’ habitats under climate change, Global
Ecology and Conservation, (2020)
doi:https://doi.org/10.1016/j.gecco.2021.e01527
This is a PDF file of an article that has undergone enhancements after acceptance,
such as the addition of a cover page and metadata, and formatting for readability,
but it is not yet the definitive version of record. This version will undergo
additional copyediting, typesetting and review before it is published in its final
form, but we are providing this version to give early visibility of the article.
Please note that, during the production process, errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier.
1
Ecological patterns and effectiveness of protected areas in the preservation of Mimusops
species’ habitats under climate change
Mimusops species’ habitat preservation under climate change
Gisèle K. S. Sinasson a,b,*, Charlie M. Shackletonb, Oscar Tekaa, Brice Sinsina
aLaboratoire d’Ecologie Appliquée, Faculté des Sciences Agronomiques, Université
d’Abomey-Calavi, 01 BP 526 Cotonou, République du Bénin; oscar_teka@yahoo.fr,
bsinsin@gmail.com
bDepartment of Environmental Science, Rhodes University, Makhanda 6140, South Africa;
c.shackleton@ru.ac.za
*Corresponding author; e-mail: sinasson.gisele@gmail.com
Abstract
Understanding the niche and habitat requirements of useful and threatened species, their shifts
under climate change and how well protected areas (PAs) preserve these habitats is relevant
for guiding sustainable management actions. Here we assessed the ecological factors
underlying the distribution of two multipurpose and threatened species, Mimusops
andongensis and M. kummel, in Benin, and potential changes in the suitable habitats covered
by PAs, under climate change scenarios. Fifty seven occurrence points were collected for M.
andongensis and 81 for M. kummel. Associations with 19 bioclimatic (from WorldClim
database) and six soil variables (from World Soil Information website) were analysed using
Principal Components Analysis, niche modelling and gap analysis. M. andongensis
occurrence is associated with high soil clay, silt, organic carbon and cation exchange capacity.
Contrastingly, M. kummel occurrence is linked to high sand content and prolonged water
holding capacity. Climatically, M. andongensis occurrence is positively related to mean
annual temperature, while M. kummel occurrence is influenced by the seasonality of
precipitation and precipitation of the wettest period. Predictions showed affinity of suitable
areas with water lines, suggesting that components of soil texture and chemical properties
should be considered during modelling. For M. andongensis suitable areas are confined to the
Guineo-Congolian zone, while for M. kummel they are mostly located in the Guineo-Sudanian
zone and absent from the driest part of the Sudanian zone. Under climate change, moderately
to highly suitable areas (probability of occurrence of species > 20 %) covered by PAs will
decrease in the case of M. andongensis, but remain stable for M. kummel. In Benin, PAs are
Journal Pre-proof
2
under threat from exploitation and uncontrolled bushfires, which may also affect populations
of the two species. Consequently, additional actions are required, including the monitoring of
species populations and the extent of different pressures, and the regularization of access to
PAs. Populations of these species outside PAs should also be given consideration because of
their very limited abundance.
Keywords: Climate change, ecological characteristics, gap analysis, habitat suitability, niche
modelling, NTFP species
1. Introduction
Timber and non-timber forest resources are important worldwide due to their
contribution to rural livelihoods and local and regional economies, as well as their
importance for ecological processes and biodiversity conservation (Chekole et al., 2015;
Swamy et al., 2018). Forest resources and non-timber forest products (NTFPs)-providing
species in particular are often confronted by multiple anthropogenic pressures, not only
because of harvesting, but also many of them are threatened by land use change, bushfires,
invasive species and herbivores (Selwood et al., 2015; Sinasson et al., 2017a; Sá et al., 2020).
Thus, despite much research and regulations, sustaining NTFP populations and benefit flows
is a substantial challenge worldwide (Wanjui, 2013; Shackleton et al., 2015). An additional
and growing threat is posed by climate change, directly through shifts in environmental
factors and indirectly by potentially increasing human dependence on forest resources
(Munang et al., 2014; You et al., 2018). All these pressures have complex and negative
impacts on populations of some species, at both species and community scales (Noulekoun et
al., 2017; Mantyka-Pringle et al., 2015). The impacts can be irreversible at the local scale for
some species (Herrero-Jauregui et al., 2013), or lead to shifts in realised and potential
distribution (Birhane et al., 2020). With possible declines in species abundances and/or
distribution comes the risk that indigenous knowledge and use practices of NTFP species
important for livelihoods, culture, domestication and safety nets, could be eroded (Chekole et
al., 2015). Consequently, in the face of the multiple and varied pressures, a comprehensive
knowledge of species ecology (i.e. species distribution and their related ecological factors)
and how that may change under climate change is necessary to underpin plausible
conservation and if necessary re-introduction strategies (Mota-Vargas and Rojas-Soto, 2012;
Mantyka-Pringle et al., 2015).
Journal Pre-proof
3
Species distribution models (SDMs) also termed Ecological Niche Models (ENMs) are
widely used to examine and define the relationship between the current occurrence of a
species within a geographical area and environmental features (Elith and Leathwick, 2009;
You et al., 2018). Thus, they help to identify the environmental variables that best predict the
current species distribution, and thereby predicting their potential future distribution under
various climate change scenarios even in regions not surveyed (Noulèkoun et al., 2017).
Though it has some limitations (Elith et al., 2011), SDM is regarded as an important tool in
conservation through identifying priority sites for particular conservation actions, in the face
of changing environmental conditions (Noulèkoun et al., 2017; Birhane et al., 2020). With
wide application, multiple methods have been developed (Elith et al., 2011), largely
prescribed by the kind of species data they use. For instance, methods such as generalized
linear models (GLMs), generalized additive models (GAMs) or boosted regression trees
(BRT) require presence and absence records (Elith et al., 2011). Such methods also work with
pseudo-absence or background data since reliable absence data are rare and difficult to obtain
(Barbet-Massin et al., 2012). However, false absence can negatively affect distribution
models and therefore to maximize the use of presence-only data, methods such as MaxEnt
have gained credence (Elith et al., 2011; Noulèkoun et al., 2017).
Protected areas (PAs) have been established worldwide to conserve ecological
processes, landscapes and populations of key resources and species (Gray et al., 2016).
Although many PAs continue to be exploited or cleared by the neighbouring inhabitants for
fields, timber and NTFPs (Geldmann et al., 2014), many PAs are important in preserving
species populations (Gray et al., 2016), especially when access is controlled. It is thus
important to analyze and understand how well the existing protected areas conserve habitats
for species of concern, to guide PA management and prioritize areas for inclusion or
establishment of new PAs. This is particularly so in the face of climate change as it alters the
suitability of current habitats in particular regions, thereby enhancing or diminishing the
conservation efficacy of PAs.
Considering the interplay of the importance of multipurpose species, PAs in sustaining
viable populations, and the potential for climate change to affect population viability or
distribution, this study assessed (i) the ecological factors associated with the distribution of
two multipurpose and threatened Mimusops species in Benin, (ii) possible changes in their
suitable habitats under climate change by 2050, and (iii) how effective the existing PAs
network is in the actual and future conservation of both species in Benin.
Mimusops andongensis Hiern and Mimusops kummel Bruce ex A. DC are two NTFP
Journal Pre-proof
4
tree species naturally occurring in several African countries, including Benin, where the two
species occur mainly in semi-deciduous and riparian forests, respectively. Both species
are exploited by local people for multiple purposes, including energy, construction, food,
alimentary uses and medicines (Lemmens et al., 2010; Soro et al., 2010; Teketay et al., 2010;
Sinasson et al., 2017b). Their fruits are also consumed by birds and other animals such as
monkeys (e.g., Nombimè and Sinsin, 2003; Kagoro-Rugunda and Hashimoto, 2015).
Furthermore, M. kummel is the most important source of pollen for honey production in
northern Benin (Tossou et al., 2011). M. andongensis has been classified as an endangered
species in Benin, according to IUCN criteria (Neuenschwander et al., 2011) but no
information exists regarding the conservation status of M. kummel in Benin. To date to
our knowledge, no efforts have been placed on understanding habitat requirements of the
species, potential impact of climate change and how well protected areas preserve these
habitats. Such investigations are important for management and conservation purposes.
2. Methods
2.1. Study area
The study was undertaken in Benin, located in the Dahomey Gap (Demenou et al.,
2018) between 6°10’ and 12°25’ north and 0°45’ and 3°55’ east. The country covers
114,763 km², with a population of approximately 10 million inhabitants in 2013 (INSAE,
2015). There is a northward climatic gradient across Benin from the Guineo-Congolian zone
in the south to the typical Sudanian zone in the north, marked by a reduction in the number of
rainy months and a decrease in mean annual rainfall and humidity (Adomou et al., 2006). In
the Guineo-Congolian (humid) zone, the climate is subequatorial with a bimodal rainfall
totalling 1,200 mm per annum (p.a). In contrast in the Sudanian (semi-arid) zone in the north
the climate is dry tropical with a unimodal rainfall, receiving less than 1,000 mm p.a. The
Guineo-Sudanian (sub-humid) zone represents the transition between these two. Benin is
divided into ten phytogeographical districts based on these climatic conditions, along with
soils types and vegetation pattern (Adomou et al., 2006). The Guineo-Congolian zone is
composed of Coast, Pobè, Ouémé Valley and Plateau phytodistricts, the Guineo-Sudanian
zone, of Zou, Bassila and South Borgou phytodistricts, and the Sudanian zone, of North
Borgou, Atacora chain and Mékrou-Pendjari phytodistricts (Fig. 1). Ecological details of the
different phytodistricts are summarized by Adomou et al. (2006). Between 1940 and 1956,
about 25 % of Benin received some sort of protected status (including 11.5 % of forest
reserves) and managed since then for the conservation of biodiversity (Sinsin et al., 2010;
Journal Pre-proof
5
Assogbadjo and Sinsin, 2010).
Figure 1. Current occurrence area of Mimusops kummel and Mimusops andongensis in Benin
2.2. Study species
M. andongensis and M. kummel belong to the Sapotaceae family. They are the two
species of the genus Mimusops naturally found in West Tropical Africa (Keay and
Hepper, 1954-1972). M. andongensis occurs in 2 (Lama and Avagbodji) out of the 16
forests in which Mimusops species have been identified in Benin (Fig. 1). Both species
were mainly exploited for medicinal purposes (e.g. to treat malaria, high blood pressure,
skin and mouth infections) but also in construction and as firewood. The most used
Journal Pre-proof
6
parts are wood, young stems, bark and leaves. Fruits are also consumed by birds and
other animals (e.g. monkeys) as well as to relieve hunger by children and during
hunting. Wood of M. kummel is sometimes exploited to make furniture and tools (e.g.
beds, chairs, wardrobes, mortars, pestles, TV poles and hoes) as well as in the
construction of bridges and roof of classrooms (Sinasson et al., 2017b).
All diameter classes were represented in the more protected forest (Lama)
relative to the 14 other forests (with missing classes). Densities of regeneration and
recruits were also higher in Lama forest than in the other forests (Sinasson, 2017).
However, in Lama forest, there is a gradient of disturbance and densities of adult trees
and regeneration of M. andongensis decreased with increasing degradation (Sinasson et
al., 2017a). For M. kummel, tree density is low in most of the forests and there is no or
very low seedlings and recruits in many forests (including protected areas). The low
seedlings and recruits in degraded sites (for M. andongensis) and in most forests (for M.
kummel), due to farm establishment, and exploitation for both timber and NTFPs, may
undermine the long-term viability of both species. Further research on seed germination
and propagation ability are necessary for plantation purposes to sustain both the uses
and populations of the species (Sinasson et al., 2017a).
2.3. Species occurrence records
Previous works have reported Mimusops species in different forests in Benin in dense
semi-deciduous and riparian forests which represent their main natural habitats (Cheke, 2001;
Adomou, 2005). Also, according to the records of Adomou (2005), M. andongensis can be
found in Plateau, Bassila, Zou and Atacora chain phytodistricts and M. kummel in Ouémé
Valley and Bassila phytodistricts. Following these authors, we prospected 28 different forests
to map the occurrence of Mimusops species in Benin. From the field survey, we identified
Mimusops species in 16 forests (with 2 forests for M. andongensis) distributed along three
bioclimatic zones. M. andongensis is found in the Guineo-Congolian zone while M. kummel
occurs in the two other climatic zones (Fig. 1). In forests with relatively low abundance of
Mimusops, all individuals were recorded using a Global Positioning System (GPS) receiver.
For forests with larger populations, we did a forest inventory (11 of the 16 forests; namely
Lama, Assanté, Savè, Idadjo, Aklamkpa, Agoua, Monts-Kouffé, Manigri, Wari-Maro,
Ouémé-Supérieur and Tanougou). Plots of 50 m x 30 m were established in the semi-
deciduous forest (Lama Forest reserve) and plots of 30 m x 30 m or 30 m x 20 m in the
riparian forests. Size of plots in riparian forests was constrained by their width. The location
Journal Pre-proof
7
of each plot was recorded via GPS as well as the presence of the species.
To increase the accuracy of habitat suitability modelling, it is recommended to consider
the presence records within the range (more extended than the study area) where the species is
influenced by similar major environmental conditions (Elith et al., 2011). For that purpose,
additional presence data of each species was obtained mainly through published papers
to cover the full occurrence range of the species. However, due to inaccuracy in these
data (Miller, 2010), we were unable to use them in our study. Indeed, records from
published papers are only a description of the species location in the form of forest or
administrative data and do not provide precise site occurrence data, especially for species with
specific habitat requirements (Guisan et al., 2006). A lack of accuracy in the analyzed data
will lead to errors in the results of the analysis, while a lack of precision can result in
conclusions of limited use (Scheldeman and van Zonneveld, 2010). In total, 57 occurrence
points were used for M. andongensis and 80 for M. kummel.
2.4. Environmental variables
For climate data, 19 bioclimatic variables were obtained from WorldClim (Version
1.4) database (http://www.worldclim.org; Hijmans et al., 2005) for both current and future
projections. For future climate conditions, we used data from the climate models HadGEM2-
ES and CNRM-CM5 based on the representative concentration pathways (RCPs) of medium
(RCP4.5) and high (RCP8.5) concentrations (Wang et al., 2019), for the horizon 2050. These
are among the most recent global climate projections by the IPCC5. Representative
Concentration Pathways (RCPs) are scenarios based on greenhouse gas concentrations
(not emissions) adopted by the IPCC5 in 2014 (Ritchie and Dowlatabadi, 2018; Wang et
al., 2019). All bioclimatic data were taken at a spatial resolution of 30s (1 km x 1 km).
Soil layers of soil types and soil characteristics (sand, silt, clay, pH, organic carbon,
CEC) were downloaded from ISRIC - World Soil Information website (Hengl et al., 2015).
We converted the layers to the 30s resolution, using the “Resample function” of Data
Management Tools in ArcGIS 10.3.
2.5. Ecological correlation with Mimusops species distribution
To identify climatic and soil factors underlying the presence of Mimusops species in
Benin, we performed a Principal Components Analysis (PCA), using R 3.1.2 (R Development
Core Team, 2017). The variables considered were soil characteristics (sand, silt, clay, pH,
organic carbon, CEC) and climatic data (annual mean temperature, annual precipitation,
Journal Pre-proof
8
precipitation of the wettest period, precipitation of the driest period and seasonality of the
precipitation). Values were extracted to the species occurrence points from the variables
layers, using Spatial Analyst Tools in ArcGIS 10.3. The first two axes explained 79.4 % of
the variance and were used to graphically display the results.
2.6. Distribution modelling and models evaluation
One of the main concerns in climate modelling using bioclimatic variables is the
correlation among variables which often leads to redundancy and can affect the accuracy of
SDMs (Phillips et al., 2006). To eliminate collinearity and non-independence of climate
dimensions (Zuur et al., 2010), we examined correlation patterns among variables to select
those that were not closely correlated. Hence, we included only a subset of variables with
Pearson correlation coefficients below 0.70 (Hipólito et al., 2015). We selected eight of the 19
bioclimatic variables, including annual mean temperature, seasonality of temperature, mean
temperature of the driest quarter, mean temperature of the warmest quarter, annual
precipitation, precipitation of the wettest period, precipitation of the driest period and
precipitation of the coldest quarter.
The ecological niche of each species was modelled using the maximum entropy
approach (MaxEnt, version 3.3.3 k; Phillips et al., 2006), which is suited to using presence-
only occurrence data (Elith et al., 2011). Models were developed maintaining the default
parameters in MaxEnt, other than using 1,000 iterations. To determine the variables that most
contributed to the model prediction, a jackknife procedure was performed. For both species,
we set 75 % of the occurrence records for the calibration of the model and the remaining 25%
were used for its evaluation. The area under the curve (AUC) of the receiver operating
characteristic (ROC) curve of the cross-validations was used to evaluate the predictive ability
of the models. The ROC curve gives a description of the relationship between the proportion
of observed presences correctly predicted (sensitivity) and the proportion of observed
absences incorrectly predicted (1- specificity) (Phillips et al., 2006). The model was then run
using the whole presence records of both species for calibration, for a better estimate of the
prediction.
2.7. Mapping and spatial gap analysis
The potential suitable habitats generated with MaxEnt, under current and future
climates, were mapped using ArcGIS 10.3. Cumulative probability (Prob.) distributions
generated by the model were used as a measure of the likelihood of occurrence of both
Journal Pre-proof
9
species. We then define five suitability levels based on the likelihood of occurrence of the
species, as follows: (i) unsuitable areas (Prob. = 0 %); (ii) poorly suitable areas (Prob. <
5 %); (iii) weakly suitable areas (Prob. 5-20 %); (iv) moderately suitable areas (Prob.
20-50 %); and (v) highly suitable areas (Prob. 50-87 %). To assess the effectiveness of the
existing PAs in the present and future conservation of Mimusops species in Benin, we
overlapped the distribution maps with the PA network in ArcGIS 10.3. Finally, we calculated
the area of predicted suitable habitats under current and future climate models, using spatial
analyst function in ArcGIS 10.3. This allowed evaluation of gain or loss in the extent of
suitable areas under climate change by 2050, and also to estimate the percentage of habitats
protected by the PA network.
3. Results
3.1. Climatic and soil factors underlying the distribution of Mimusops species in Benin
The correlation analysis among the ecological variables and the two axes separated
two different groups. One was M. andongensis occurrence with the annual mean temperature,
clay, silt, CEC and OC content of the soil at the positive side of the first axis. The other was
M. kummel occurrence with the precipitation of the wettest period, the seasonality of
precipitation and the sand content of the soil at the negative side. The second axis opposed
annual precipitation and the precipitation of the driest period at the positive side, and the soil
pH at the negative side (Fig. 2). These variables discriminated by the second axis were weakly
(r < 0.5) correlated with Mimusops species occurrence.
Journal Pre-proof
10
Figure 2. Principal Components Analysis results on Mimusops species occurrence and
ecological factors. MK = M. kummel, MA = M. andongensis, OC = Organic carbon, Bio1 = Mean annual temperature,
Bio12 = Annual precipitation, Bio13 = Precipitation of the wettest period, Bio14 = Precipitation of the driest period, Bio15 =
Precipitation seasonality.
3.2. Current potential suitable ecological niche and changes under climate change
The AUC values obtained from the jackknife test indicated all the models (with
current climate and future climate projections under HadGEM2-ES and CNRM-CM5)
had a good predictive ability. For M. andongensis, the AUC was greater than 0.99 during both
the test and the final training. In the case of M. kummel, AUC was 0.9 during the test and 0.96
(or greater) for the final training. However, not all the variables considered contributed greatly
(Table 1). CEC, clay and sand content of the soil, and precipitation of the coldest quarter
contributed most to M. andongensis models while the soil type and pH of the soil contributed
less, under both RCP8.5 and RCP4.5 scenarios. For M. kummel, mean temperature of the
driest quarter, precipitation of the coldest quarter, seasonality of temperature and precipitation
of the wettest period contributed more to the obtained models while the soil type, clay content
and pH of the soil contributed the least (Table 1).
Table 1. Percentage contribution (%) of listed variables to Mimusops species models
Variables
M. andongensis
M. kummel
RCP8.5
RCP4.5
RCP8.5
RCP4.5
Annual Mean Temperature
26.9
17.3
15.8
15.3
Annual Precipitation
13.5
9.6
7.9
8.2
Journal Pre-proof
11
Mean Temperature of the Driest Quarter
29.8
39.4
38.0
28.4
Mean Temperature of the Warmest Quarter
23.1
14.4
21.6
27.3
Precipitation of the Coldest Quarter
54.9
41.8
36.8
31.0
Precipitation of the Driest Period
25.0
25.6
26.3
25.1
Precipitation of the Wettest Period
37.5
36.1
31.6
30.9
Seasonality of the Temperature
37.5
34.8
35.1
29.7
CEC of the soil
81.6
83.0
29.1
21.5
Clay content of the soil
65.5
64.8
5.7
6.2
Organic Carbon content of the soil
41.7
40.1
23.9
20.3
pH of the soil
3.9
2.3
6.1
6.3
Sand content of the soil
53.8
53.3
8.7
9.1
Silt content of the soil
12.2
9.8
13.4
15.8
Type of soil
2.0
1.0
5.3
8.6
For M. andongensis, moderately to highly suitable areas (probability of occurrence >
20 %) were mainly confined to the Guineo-Congolian zone of the country, under both current
and future climates. Results showed an increase in the extent (25.6 % and 9.9 % under
CNRM-CM5 and HadGEM2-ES climate models, respectively) and the level of suitability of
favourable habitats (Fig. 3) with the scenario RCP8.5, compared to the current situation.
However, there was a decrease (16.7 % and 26.7 % under CNRM-CM5 and HadGEM2-ES
climate models, respectively) in the highly suitable habitats (probability of occurrence > 50
%). With the scenario RCP4.5, there was an increase in suitable habitats (32.5 %) under
CNRM-CM5 model while a decrease (4.6 %) was observed under HadGEM2-ES model.
Similarly to the prediction under the scenario RCP8.5, there was a decrease (9.5 % and
14.3 % under CNRM-CM5 and HadGEM2-ES climate models, respectively) in the
highly suitable habitats (probability of occurrence > 50 %) with the scenario RCP4.5.
All the models also predicted poorly suitable areas (probability of occurrence < 5 %) in the
Mékrou-Pendjari phytodistrict, which will slightly increase in extent in the future (Fig. 3).
In the case of M. kummel, suitable habitats were mainly located in the Guineo-
Sudanian and the Sudanian zones (but mostly in the former), except the far north and driest
part of the country. The most suitable area for M. kummel in the Sudanian zone was located in
the North-west of that zone, in the Atacora chain and Mékrou-Pendjari phytodistricts (Fig. 4).
There was an increase in the extent (5.9 % and 9.4 % under CNRM-CM5 and HadGEM2-ES
climate models, respectively) of suitable habitats with the scenario RCP8.5, compared to the
current projection. Also, highly suitable areas increased of 5.1 % under CNRM-CM5, while
there was a slight decrease of 1 % under HadGEM2-ES. Similarly to the prediction under
the scenario RCP8.5, there was an increase in the extent (7.2 % and 22.4 % under
CNRM-CM5 and HadGEM2-ES climate models, respectively) of suitable habitats with
Journal Pre-proof
12
the scenario RCP4.5. Also in terms of the highly suitable habitats (probability of
occurrence > 50 %), there was an increase in their extent (9.8 %) under CNRM-CM5
model while a decrease (7.2 %) was observed under HadGEM2-ES model. The models
also predicted favourable habitats (probability of occurrence < 20 %) in the Guineo-
Congolian zone, with an increase in the future mainly in the Coast phytodistrict.
For M. kummel, suitable habitats showed a tendency to align with water drainage
systems. The same trend has been observed for M. andongensis, beyond Lama Forest reserve.
For both species, the occurrence records used for the modelling were well covered by the
predicted suitable areas with a probability of occurrence higher than 20 %.
A
C
B
Journal Pre-proof
13
Figure 3. Distribution predictions of Mimusops andongensis under current (A) and future
projections in 2050 under (B) CNRM-CM5 and (C) HadGEM2-ES model climates with
scenario RCP8.5, and (D) CNRM-CM5 and (E) HadGEM2-ES model climates with scenario
RCP4.5
D
E
Journal Pre-proof
14
C
B
A
Journal Pre-proof
15
Figure 4. Distribution predictions of Mimusops kummel under current (A) and future
projections in 2050 under (B) CNRM-CM5 and (C) HadGEM2-ES model climates with
scenario RCP8.5, and (D) CNRM-CM5 and (E) HadGEM2-ES model climates with scenario
RCP4.5
3.3. Effectiveness of protected areas in conserving Mimusops species under current and
future climate
According to the gap analysis, protected areas covered 21 % of suitable areas for M.
andongensis, which will increase to 27 % under future climate. However, only one protected
area (Lama Forest reserve) protects and will continue to protect the species. Indeed, this
protected area covered 31 % of suitable habitats with a probability of occurrence higher than
20 %, but this proportion will decrease to 24 % in the future (Fig. 5).
In the case of M. kummel, 20 % of the current favourable range occurred with PAs and
this included 28 % of suitable areas with a probability of occurrence higher than 20 %.
Opposite to M. andongensis, the extent of suitable habitats currently covered by PAs will
decrease to 17 % in the future but the extent of suitable areas with a probability of occurrence
higher than 20 % will remain stable (Fig. 6). In the Guineo-Sudanian zone, almost all the
D
E
Journal Pre-proof
16
existing PAs cover a part of the suitable areas whereas in the Sudanian zone, the more suitable
areas (probability of occurrence > 20 %) are outside of PAs; only a small portion of suitable
areas with probability of occurrence less than 20 % is covered by the Pendjari National Park,
especially the hunting zone.
C
B
A
Journal Pre-proof
17
Figure 5. The portion of Mimusops andongensis suitable areas covered by protected areas,
under current (A) and future projections in 2050 under (B) CNRM-CM5 and (C) HadGEM2-
ES model climates with scenario RCP8.5, and (D) CNRM-CM5 and (E) HadGEM2-ES model
climates with scenario RCP4.5
D
E
Journal Pre-proof
18
C
B
A
Journal Pre-proof
19
Figure 6. The portion of Mimusops kummel suitable areas covered by protected areas, under
current (A) and future projections in 2050 under (B) CNRM-CM5 and (C) HadGEM2-ES
model climates with scenario RCP8.5, and (D) CNRM-CM5 and (E) HadGEM2-ES model
climates with scenario RCP4.5
4. Discussion
Knowledge of the distribution area of specific species, together with the underlying
correlated bioclimatic factors is fundamental for decision-making in conservation (Mota-
Vargas and Rojas-Soto, 2012; Hipólito et al., 2015). An inaccurate mapping of a species
distribution and ecology can misdirect conservation measures. Contrary, a reliable one can
guide conservation measures, including domestication plans, if needed, by identifying
appropriate sites for introduction. This is very more important in the face of climate change
which is altering species distributions and thereby reshaping the efficacy of PA networks and
management plans (Feeley and Silman, 2016).
The analysis indicated that M. andongensis occurs on fertile soils characterized by
high clay, silt, CEC and OC, while M. kummel occurrence is associated with high sand
content and moisture. This is confirmed by M. kummel being common along watercourses in
sub-humid and semi-arid zones, while M. andongensis occurs on soils with a more compact
structure and a high water retention capacity because of high clay content, in the humid zone
with two rainy seasons. The preference of both Mimusops species for moist conditions has
been previously noted (Lemmens, 2005; Bein et al., 1996). The presence of M. andongensis is
D
E
Journal Pre-proof
20
also positively correlated with the mean annual temperature, evidenced by the restriction of
the species to the humid zone and thus fluctuations in the temperature could be a limiting
factor. The distribution of M. kummel is positively correlated with the seasonality of the
precipitation and the precipitation of the wettest period, indicating its preference for wetter
sites within the broader less humid climatic zones (Guineo-Congolian and Sudanian) which
are characterized by one rainy season (Adomou, 2010). Results showed that pH was not a
factor influencing the distribution of Mimusops species in Benin.
The occurrence records used for the modelling of both species were well covered by
the predicted suitable areas with a probability of occurrence higher than 20 %. This suggests,
apart from the high values (> 0.9) of AUC obtained for the models, that the models used have
a good predictive ability. This is because sites of good probability of occurrence covered all
presence points, while areas with few or no presence records displayed null or low probability
of occurrence (Saupe et al., 2012).
The models projected suitable habitats in the moister parts of the country for both
species. For instance, for M. andongensis, suitable areas with a probability of occurrence
more than 20 % were located in the Guineo-Congolian zone which is the most humid part of
the country under both current and future climates with both RCP8.5 and RCP4.5 scenarios.
For M. kummel, suitable habitats were mostly located in the Guineo-Sudanian (sub-humid)
zone. Also, in the Sudanian zone, suitable areas are absent from the far north and driest part of
the country (Adomou et al., 2006). Moreover, the species displayed an affinity with water
lines. For instance, suitable habitats for M. kummel followed the water lines system and
similar trend has been observed for M. andongensis, beyond Lama Forest reserve. This
confirms that both species have preference for less dry conditions, and occur on soils that
maintain some moisture during the year (Ake-Assi, 2001).
Although soil type has been found to be an important variable in determining the
ecological niche of many plant species (Miller, 2010; Birhane et al., 2020), results from this
study showed that taking into account soil microvariation could give better results. For
instance, for M. andongensis, the variable “soil type” contributed less to the models,
compared to the soil clay and sand content, cation exchange capacity (CEC) and organic
carbon (OC). Besides, the variable “soil type” contributed the least in predicting the
distribution of the study species and a similar result has been obtained for Adansonia
digitata L. in Ethiopia (Birhane et al., 2020). Furthermore, the more suitable (moderately
to highly suitable) areas, for both species, tend to align with water drainage systems; this
might be due to the fact that we used sand, silt, clay and OC contents of the soil as well
Journal Pre-proof
21
as soil pH and CEC as input variables for the modelling, and not only the variable “soil
type”. Thus, although it is very complex to identify all the factors affecting species
distribution and viability (Araujo and Guisan, 2006), it is important to consider, together with
the variable “soil type”, the components of soil texture and chemical properties associated
with species occurrence while modelling their ecological niche. This may give more precise
results (Liu et al., 2018). However, other factors such as competition, dispersal limitation and
human disturbance are important to consider in the modelling as they may have influenced the
species current distribution (Miller, 2010; Mockta et al., 2018).
The results showed an increase in the extent and the level of suitability of favourable
habitats for both species under both future climate models, compared to the current situation.
However, there was a decrease in the highly suitable habitats (probability of occurrence > 50
%) under future climate for M. andongensis suggesting that the species will be more
threatened as it already has a very limited distribution and suitable habitats range. A decrease
was also predicted in the extent of highly suitable areas of M. kummel, but only under one of
the two climate models considered (HadGEM2-ES). However, opposite to M. andongensis,
M. kummel has a larger suitable habitat range and the decrease rate is estimated at 1 %. Our
findings corroborate with previous works on the impact of climate change on habitat
suitability. For instance, stability was mainly predicted in the distribution of eight palm
species in Benin (Idohou et al., 2017) whilst decrease in suitable habitats has been mainly
predicted for Adansonia digitata L. and Faidherbia albida (Delile) A. Chev. in Ethiopia
(Noulèkoun et al., 2017; Birhane et al., 2020).
Globally, similar trend was observed in the change in the extent of favourable
habitats for both species under future climate models for both scenarios, compared to
the current situation. Also, results confirmed the impact of climate to be less under the
scenario of lower emission (RCP4.5) than under the one of high emission (RCP8.5)
(Moore et al., 2013). For instance, increase in the extent of suitable habitats under
RCP4.5 was higher than under RCP8.5. Similarly, decrease in the extent of highly
suitable habitats under RCP4.5 was lower than under RCP8.5. Furthermore, the same
climatic variables appeared to be more influential on Mimusops species’ distribution
under both scenarios, although not in the same order of importance. These are the
precipitation of the coldest quarter, mean temperature of the driest quarter, seasonality
of temperature and precipitation of the wettest period. Extracted values of these
variables for both species globally showed increase by horizon 2050 under both
scenarios, compared to historical values. Similar increase in mean values of
Journal Pre-proof
22
precipitation and temperature was highlighted by Wang et al. (2019). However,
increases in precipitations are higher under RCP4.5 while increases in temperatures are
higher under RCP8.5. This might be the cause of increase in globally suitable habitats
being higher under RCP4.5 than under RCP8.5 while decrease in highly suitable
habitats was higher under RCP8.5.
According to the gap analysis, only one PA (Lama Forest reserve) protects and will
continue to protect M. andongensis. Indeed, it is the only PA that covers suitable sites with a
probability of occurrence of more than 20 %, and harbours 31 % of these suitable habitats.
However, this proportion will decrease to 24 % in the future. M. andongensis is protected by
Lama Forest reserve and as the most secure of existing PAs in Benin, threat levels are
reasonably low. However, Lama Forest reserve suffers from episodic, anthropogenic fires
(ONAB, 2011) which may pose some threat, but the effects of these fires on M. andongensis
populations still need to be investigated. For M. kummel, 20 % of the favourable range was
within the PAs network, including 28 % of suitable areas with a probability of occurrence
higher than 20 %. Although most of suitable areas are outside PAs, a good portion is covered
by and distributed among many PAs, especially in the Guineo-Sudanian zone. Thus, if PAs
are well maintained, M. kummel could be well protected but this will require monitoring.
However, this is not currently the case for many PAs in Benin (Houehanou et al., 2013) as
most PAs are still accessed or exploited by local people, as also occurs in other countries
around the world (Geldmann et al., 2014; Gray et al., 2016; Sarathchandra et al., 2018).
This may undermine the effectiveness of PAs in their role of protection of threatened and
important species. For instance, some trees of M. kummel selected and marked for the
monitoring as part of this research were cut by local people (G. Sinasson, unpubl. data).
Knowing the effectiveness of PAs in preserving target species populations could help
in planning future actions. For instance, if species are not well covered by the existing PAs,
based on their importance, distribution and abundance, it could be motivation for declaring of
another PA (Gray et al., 2016) or sensitizing neighbouring communities on strategies of
harvesting that can limit impacts on species populations. The latter requires the identification
and understanding of such strategies, especially endogenous ones.
5. Conclusion
This study showed that M. andongensis and M. kummel have preference for moist
conditions and occur on soils with high soil moisture throughout most of the year. This
indicates that climatic conditions with a shorter rainy season and consequently more dry
months could negatively affect populations of both species. Our results suggest that it is
Journal Pre-proof
23
important to consider components of soil texture and chemical properties, not just soil types,
for distribution modelling of species as it can increase the precision of the predictions.
Favourable habitats for M. andongensis are confined to the Guineo-Congolian zone, while for
M. kummel, they are mostly located in the Guineo-Sudanian zone and absent from the driest
part of the Sudanian zone. Also, M. andongensis has a very limited distribution in Benin
compared to M. kummel confirming the conservation status attributed to M. andongensis.
According to the current distribution status of both species, they could be well preserved by
the existing PAs. However, because all PAs in Benin are still under exploitation or impact of
fires, the protection of individual species cannot be guaranteed. Moreover, predictions showed
a decrease in suitable habitats for M. andongensis covered by PAs (only Lama Forest
reserve), probably hindering any re-introduction opportunities for the species. However, the
extent of suitable areas for M. kummel inside PAs remained stable under future climate
models. Thus, conservation actions should mainly be focussed on M. andongensis due to its
limited distribution and already threatened status. Such actions include (1) the monitoring of
existing populations to assess impacts of different pressures, (2) the development of suitable
sylvicultural knowledge and practices for introduction purposes, and (3) the ex-situ
conservation of the species.
Acknowledgements
Huge thanks go to Cyrus Binassoua, Christian Affoukou, Hervé Kanlissou, Michaël Hounsa
and local guides in the different sites for field assistance.
Funding
This work was supported by the International Foundation for Science (grant D/5467-1, 2013);
the Organization for Women in Science for the Developing World and Swedish International
Development Cooperation Agency (grant no. 3240266463, 2013). IDEA WILD also provided
some fieldwork materials. The contribution of CS was supported by the South African
Research Chairs Initiative of the Department of Science and Innovation and the National
Research Foundation of South Africa (grant no. 84379). Any opinion, finding, conclusion or
recommendation expressed in this material is that of the authors and the NRF does not accept
any liability in this regard.
References
Adomou, A.C. 2005. Vegetation patterns and environmental gradients in Benin:
implications for biogeography and conservation. PhD dissertation, Wageningen University,
Wageningen, Netherlands; https://edepot.wur.nl/121707.
Journal Pre-proof
24
Adomou, C.A., Sinsin, B., and van der Maesen, L.J.G. 2006. Phytosociological and
chorological approaches to phytogeography: a meso-scale study in Benin. Systematics and
Geography of Plants 76: 155-178; https://www.jstor.org/stable/20649708.
Adomou, A. 2010. Territoires phytogéographiques au Bénin. In Sinsin B. &
Kampmann D. (eds). 2010: Biodiversity Atlas of West Africa, Volume I: Benin. Cotonou &
Frankfurt/Main, pp 120-143; https://www.uni-frankfurt.de/47671063/BJ-Atlas.
Aké-Assi, L. 2001. Observations sur la diversité des Sapotaceae de la flore naturelle
de la Côte d'Ivoire. Systematics and Geography of Plants 71: 187-195; doi: 10.2307/3668666.
Araujo, M.B., and Guisan, A. 2006. Five (or so) challenges for species distribution
modelling. Journal of Biogeography 33: 1677-1688; doi: 10.1111/j.1365-2699.2006.01584.x.
Assogbadjo, A.E., and Sinsin, B. 2010. Zones of special ecological importance for
biodiversity conservation in Benin: state forest reserves. In Sinsin B. & Kampmann D. (eds).
2010: Biodiversity Atlas of West Africa, Volume I: Benin. Cotonou & Frankfurt/Main, pp
530-535; https://www.uni-frankfurt.de/47671063/BJ-Atlas.
Barbet-Massin, M., Jiguet, F., Albert, C.H., and Thuiller, W. 2012. Selecting pseudo-
absences for species distribution models: how, where and how many? Methods in Ecology
and Evolution 3: 327-338; doi: 10.1111/j.2041-210X.2011.00172.x.
Bein, E., Habte, B., Jaber, A., Birnie, A., and Tengnas, B. 1996. Useful Trees and
Shrubs in Eritrea: Identification, Propagation and Management for Agricultural and Pastoral
Communities. Technical Handbook 12, Regional Soil Conservation Unit, RSCU/SIDA
Nairobi, Kenya.
Birhane, E., Asgedom, K.T., Tadesse, T., Hishe, H., Abrha, H., and Noulèkoun, F.
2020. Vulnerability of baobab (Adansonia digitata L.) to human disturbances and climate
change in western Tigray, Ethiopia: Conservation concerns and priorities. Global Ecology and
Conservation 22, e00943; doi: 10.1016/j.gecco.2020.e00943.
Cheke, R.A. 2001. Benin. In Fishpool L.D.C. & Evans M. I. (eds). 2001. Important
Bird Areas in Africa and associated islands: Priority sites for conservation. Newbury and
Cambridge, UK: Pisces Publications and BirdLife International (BirdLife Conservation Series
No. 11), pp 93-98; http://datazone.birdlife.org/userfiles/file/IBAs/AfricaCntryPDFs/Benin.pdf
Chekole, G., Asfaw, Z., and Kelbessa, E. 2015. Ethnobotanical Study of Medicinal
Plants in the Environs of Tara-gedam and Amba Remnant Forests of Libo Kemkem District,
Northwest Ethiopia. Journal of Ethnobiology and Ethnomedicine 11: 1-38; doi:
10.1186/1746-4269-11-4.
Journal Pre-proof
25
Demenou, B.B., Doucet, J-L., and Hardy, O.J. 2018. History of the fragmentation of
the African rain forest in the Dahomey Gap: insight from the demographic history of
Terminalia superba. Heredity 120: 547-561; doi: 10.1038/s41437-017-0035-0.
Elith, J., and Leathwick, J.R. 2009. The contribution of species distribution modelling
to conservation prioritization. Spatial Conservation Prioritization: Quantitative Methods &
Computational Tools (ed. by A. Moilanen, K.A. Wilson and H.P. Possingham), pp. 70-93,
Oxford University Press, Oxford, UK.
Elith, J., Phillips, S.J., Hastie, T., Dudik, M., Chee, Y.E. and Yates, C.J. 2011. A
statistical explanation of MaxEnt for ecologists. Diversity Distribution 17: 43-57; doi:
10.1111/j.1472-4642.2010.00725.x.
Feeley, K.J., and Silman, M.R. 2016. Sustainability of culturally important teepee
poles on Mescalero Apache Tribal Lands: Characteristics and climate change effects.
Diversity and Distributions 22: 1081-1084; doi: 10.1111/ddi.12475.
Gray, C., Hill, S., Newbold, T. et al. 2016. Local biodiversity is higher inside than
outside terrestrial protected areas worldwide. Nature Communications 7, 12306; doi:
10.1038/ncomms12306.
Guisan, A., Broennimann, O., Engler, R., Vust, M., Yoccoz, N.G., Lehmann, A., and
Zimmermann, N.E. 2006. Using niche-based distribution models to improve the sampling of
rare species. Conservation Biology 20: 501-511; doi: 10.1111/j.1523-1739.2006.00354.x.
Hengl, T., Heuvelink, G.B.M., Kempen, B., Leenaars, J.G.B., Walsh, M.G., Shepherd,
K.D., et al. 2015. Mapping Soil Properties of Africa at 250 m Resolution: Random Forests
Significantly Improve Current Predictions. PLoS ONE 10: e0125814; doi:
10.1371/journal.pone.0125814.
Herrero-Jauregui, C., Guariguata, M.R., Cardenasm, D., Vilanova, E., Robles, M.,
Licona, J.C., and Nalvarte, W. 2013. Assessing the Extent of “Conflict of Use” in
Multipurpose Tropical Forest Trees: A Regional View. Journal of Environmental
Management 130: 40-47.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., and Jarvis, A. 2005. Very high
resolution interpolated climate surfaces for global land areas. International Journal of
Climatology 25: 1965-1978; doi: 10.1002/joc.1276.
Hipólito, J., Hasui, E., and Viana, B.F. 2015. Solving problems involving the
distribution of a species of unknown distribution via ecological niche modeling. Research
Letters, Natureza & Conservação 13: 15-23; doi: 10.1016/j.ncon.2015.03.001.
Journal Pre-proof
26
Houehanou, T.D., Assogbadjo, A.E., Glèlè Kakaï, R., Kyndt, T., Houinato, M., and
Sinsin, B. 2013. How far a protected area contributes to conserve habitat species composition
and population structure of endangered African tree species (Benin, West Africa). Ecological
Complexity 13: 60-68; doi: 10.1016/j.ecocom.2013.01.002.
Idohou, R., Assogbadjo, A.E., Glèlè Kakaï, R., and Peterson, A.T. 2017. Spatio-
temporal dynamic of suitable areas for species conservation in West Africa: eight
economically important wild palms under present and future climates. Agroforestry Systems,
91: 527-540; doi: 10.1007/s10457-016-9955-6.
INSAE (National Institute of Statistics and Economic Analysis). 2015. RGPH4: Que
Retenir des effectifs de population en 2013? Direction des études démographiques, Ministère
du Développement, de l’Analyse économique et de la Prospective, Bénin. 33p.
Kagoro-Rugunda, G., and Hashimoto, C. 2015. Fruit phenology of tree species and
chimpanzees’ choice of consumption in Kalinzu Forest Reserve, Uganda. Open Journal of
Ecology 5: 477-490; doi: 10.4236/oje.2015.510039.
Keay, R.W.J., and Hepper, F.N. (eds.), 1954-1972. Flora of West Tropical Africa by
Hutchinson J and Dalziel JM, 2nd edition Vol. I (828 pp.), Vol. II (544 pp.), Vol. III (574
pp.). Millbank, London.
Lemmens, R.H.M.J. 2005. Mimusops andongensis Hiern / Mimusops kummel Bruce ex
A. DC. In D. Louppe, A.A. Oteng-Amoako and M. Brink (eds.). Plant Resources of Tropical
Africa/Ressources Végétales de l’Afrique Tropicale. Wageningen, Netherlands.
http://www.prota4u.org/search.asp.
Lemmens, R. H. M. J., Omino, E. A., and Bosch, C. H. (Eds.). 2010. Bois d’oeuvre de
l’Afrique tropicale. Conclusions et recommandations basées sur PROTA 7(1) : “Bois
d’oeuvre 1”. [Traduction de : Timbers of Tropical Africa. Conclusions and recommendations
based on PROTA 7(1): ‘Timbers 1’. 2009]. Fondation PROTA, Nairobi, Kenya, 92 pp.
Liu, C., Chen, L., Tang, W., Peng, S., Li, M., Deng, N., and Chen, Y. 2018. Predicting
Potential Distribution and Evaluating Suitable Soil Condition of Oil Tea Camellia in China.
Forests 9: 487; doi: 10.3390/f9080487.
Mantyka-Pringle, C.S., Visconti, P., Di Marco, M., Martin, T.G., Rondinini, C., and
Rhodes, J.R., 2015. Climate change modifies risk of global biodiversity loss due to land-cover
change. Biological Conservation 187: 103-111; doi: 10.1016/j.biocon.2015.04.016.
Miller, J. 2010. Species Distribution Modeling. Geography Compass 4: 490-509; doi:
10.1111/j.1749-8198.2010.00351.x.
Journal Pre-proof
27
Mockta, T.K., Fulé, P.Z., Meador, A.S., Padilla, T., and Kim, Y. 2018. Sustainability
of culturally important teepee poles on Mescalero Apache Tribal Lands: Characteristics and
climate change effects. Forest Ecology and Management 430: 250-258; doi:
10.1016/j.foreco.2018.08.017.
Moore, J.K., Lindsay, K., Doney, S.C., Long, M.C., and Misumi, K. 2013. Marine
Ecosystem Dynamics and Biogeochemical Cycling in the Community Earth System Model
[CESM1(BGC)]: Comparison of the 1990s with the 2090s under the RCP4.5 and RCP8.5
Scenarios. Journal of Climate 26: 9291-9312; doi: 10.1175/JCLI-D-12-00566.1.
Mota-Vargas, C., and Rojas-Soto, O.R. 2012. The importance of defining the
geographic distribution of species for conservation: The case of the Bearded Wood-Partridge.
Journal for Nature Conservation 20: 10-17; doi: 10.1016/j.jnc.2011.07.002.
Munang, R., Andrews, J., Alverson, K., and Mebratu, D. 2014. Harnessing
Ecosystem-based Adaptation to Address the Social Dimensions of Climate Change,
Environment: Science and Policy for Sustainable Development, 56: 18-24; doi:
10.1080/00139157.2014.861676.
Neuenschwander, P., Sinsin, B., and Goergen, G. (eds.), 2011. Nature Conservation in
West Africa: Red List for Benin. Protection de la Nature en Afrique de l’Ouest: Une liste
Rouge pour le Bénin. International Institute of Tropical Agriculture, Ibadan, Nigeria, 365p.
Nombimè, G., and Sinsin, B. 2003. Les stratégies de survie du singe à ventre rouge
(Cercopithecus erythrogaster erythrogaster) dans la Forêt Classée de la Lama.
Biogeographica 79: 153-166.
Noulekoun, F., Birhane, E., Chude, S., and Zenebe, A. 2017. Characterization
of Faidherbia albida (Del.) A. Chev. population in agroforestry parklands in the highlands of
Northern Ethiopia: impact of conservation, environmental factors and human
disturbances. Agroforest Syst 91: 123-135; doi: 10.1007/s10457-016-9910-6.
Noulèkoun, F., Chude, S., Zenebe, A., and Birhane, E. 2017. Climate change impacts
on Faidherbia albida (delile) A. Chev. distribution in dry lands of Ethiopia. African Journal
of Ecology 55: 233-243; doi: 10.1111/aje.12345.
ONAB (National Office of Wood) 2011. Plan Directeur d’Aménagement et de Gestion
du Noyau Central de la Forêt Classée de la Lama 2011-2020. Ministère de l’Environnement et
de la Protection de la Nature, Cotonou, Bénin. 98p.
Phillips, S.J., Anderson, R.P., and Schapire, R.E. 2006. Maximum entropy modelling
of species geographic distributions. Ecological Modelling 190: 231-259; doi:
10.1016/j.ecolmodel.2005.03.026.
Journal Pre-proof
28
R Core Team 2017. R: a language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria; http://www.R-project.org/.
Ritchie, J., and Dowlatabadi, H. 2018. Defining climate change scenario
characteristics with a phase space of cumulative primary energy and carbon intensity.
Environmental Research Letters 13: 024012; doi: 10.1088/1748-9326/aaa494.
Sá, D., Scariot, A., and Ferreira, J.B. 2020. Effects of ecological and anthropogenic
factors on population demography of the harvested Butia capitatapalm in the Brazilian
Cerrado. Biodiversity and Conservation 29: 1571-1588; doi: 10.1007/s10531-018-1669-9.
Sarathchandra C., Dossa, G.G.O., Ranjitkar, N.B., et al. 2018. Effectiveness of
protected areas in preventing rubber expansion and deforestation in Xishuangbanna,
Southwest China. Land Degradation and Development 29: 2417-2427; doi: 10.1002/ldr.2970.
Saupe, E.E., Barve, V., Myers, C.E., Soberon, J., Barve, N., Hensz, C.M., and Lira-
Noriega, A. 2012. Variation in niche and distribution model performance: the need for a priori
assessment of key causal factors. Ecological Modelling 237: 11-22; doi:
10.1016/j.ecolmodel.2012.04.001.
Scheldeman, X., and Van Zonneveld, M. 2010. Training Manual on Spatial Analysis
of Plant Diversity and Distribution. Bioversity International, Rome, Italy. Retrieved from
http://www.bioversityinternational.org/uploads/tx_news/Training_manual_on_spatial_analysi
s of_plant_diversity_and_distribution_1431_07.pdf.
Selwood, K.E., McGeoch, M.A., and Mac Nally, R. 2015. The effects of climate
change and land-use change on demographic rates and population viability. Biological
Reviews 90: 837-853; doi: 10.1111/brv.12136.
Shackleton, C.M., Pandey, A.K., and Ticktin, T. 2015. Ecologically sustainable
harvesting of non-timber forest products: Disarming the narrative and the complexity. In
Shackleton C.M. et al. (eds). 2015. Ecological Sustainability for Non-timber Forest Products:
Dynamics and Case Studies of Harvesting. Taylor & Francis Group, pp 260-278.
Sinasson S., G.K. 2017. Ethnobotany and ecology of Mimusops andongensis Hiern
and Mimusops kummel Bruce ex A. DC: implications for the species management and
conservation in Benin (West Africa). PhD dissertation, Faculty of Agronomic Sciences,
University of Abomey-Calavi (FSA/UAC), Abomey-Calavi, Benin.
Sinasson S., G.K, Shackleton, C.M., Glèlè Kakaï, R.L., and Sinsin, B. 2017a. Forest
degradation and invasive species synergistically impact Mimusops andongensis (Sapotaceae)
in Lama Forest Reserve. Biotropica 49: 160-169; doi: 10.1111/btp.12370.
Journal Pre-proof
29
Sinasson S., G.K, Shackleton, C.M., Assogbadjo, A.E., and Sinsin, B. 2017b. Local
Knowledge on the Uses, Habitat and Change in Abundance of Multipurpose Mimusops
Species in Benin. Economic Botany 71: 105-122; doi: 10.1007/s12231-017-9370-6.
Sinsin, B., Kassa, B., and Kidjo, F. 2010. Zones of special ecological importance for
biodiversity conservation in Benin: network of protected areas of Benin. In Sinsin B. &
Kampmann D. (eds). 2010: Biodiversity Atlas of West Africa, Volume I: Benin. Cotonou &
Frankfurt/Main, pp 522-529; https://www.uni-frankfurt.de/47671063/BJ-Atlas.
Soro, D., Kone, M.W., and Kamanzi, K. 2010. Evaluation des Activités
Antimicrobiennes et Anti-Radicaux Libres de Quelques Taxons Bioactifs de Côte D’ivoire.
European Journal of Scientific Research 40: 307-317.
Swamy, L., Drazen, E., Johnson, W.R., and Bukoski, J.J. 2018. The future of tropical
forests under the United Nations Sustainable Development Goals. Journal of Sustainable
Forestry 37: 221-256; doi: 10.1080/10549811.2017.1416477.
Teketay, D., Senbeta, F., Maclachlan, M., Bekele, M., and Barklund, P. 2010. Edible
wild plants in Ethiopia. Addis Ababa University Press, Addis Ababa, Ethiopia, 575 pages.
Tossou, G. M., Yédomonhan, H., Adomou, A.C., Demènou, B. B., Akoègninou, A., et
Traoré, D. 2011. Caracterisation pollinique des miels d’un élevage apicole dans
l’arrondissement de Manigri en zone soudano-guinéenne au Bénin. Annale Botanique de
l’Afrique de l’Ouest 7: 42-58.
Wang, J., Lin, H., Huang, J., Jiang, C., Xie, Y., and Zhou, M. 2019. Variations of
Drought Tendency, Frequency, and Characteristics and Their Responses to Climate Change
under CMIP5 RCP Scenarios in Huai River Basin, China. Water 11: 2174; doi:
10.3390/w11102174.
Wanjui, J. 2013. Biodiversity Conservation Needs and Method to Conserve the
Biological Diversity. Journal of Biodiversity and Endangered Species 1: 113; doi:
10.4172/2332-2543.1000113.
You, J., Qin, X., Ranjitkar, S. et al. 2018. Response to climate change of montane
herbaceous plants in the genus Rhodiola predicted by ecological niche modelling. Scientific
Report 8: 5879; doi: 10.1038/s41598-018-24360-9.
Zuur, A.F., Ieno, E.N., and Elphick, C.S. 2010. A protocol for data exploration to
avoid common statistical problems. Methods in Ecology and Evolution 1: 3-14; doi:
10.1111/j.2041-210X.2009.00001.x.
Journal Pre-proof
30
Highlights
We assessed two Mimusops species protection by protected areas (PAs) under
current/future climate
Components of soil texture and chemical properties are important for niche modelling
Suitable areas under PAs will reduce for M. andongensis and be stable for M. kummel
Additional actions (e.g. population monitoring) will help to preserve the species
Journal Pre-proof
31
Declaration of interests
xThe authors declare that they have no known competing financialinterestsor personal relationships
that could have appeared to influence the work reported in this paper.
☐The authors declare the following financial interests/personal relationships which may be
considered as potential competing interests:
Journal Pre-proof
