Research Paper

INTRODUCTION

Bird populations are declining globally (Gaston et al. 2003, Spooner et al. 2018, Bowler et al. 2019). This unfortunate trend is also evident among African birds (Thiollay 2006, BirdLife International 2013, Cooper et al. 2017, Beresford et al. 2019), posing particular concern for species with low population sizes and those living in endemic regions such as the mountains of West Africa (De Klerk et al. 2002). Even though relatively little information exists on the biodiversity of West African mountains (Hořák et al. 2023), available studies report on avian population declines in West Africa and Cameroon, regions otherwise known as biodiversity hotpots (Fjeldsaå et al. 1997, Whytock et al. 2018, Riegert et al. 2021).

As geographical distribution in birds is strongly determined by their habitat preferences (Ricklefs, 2013) and habitat deterioration is one of the major causes of avian population declines (Dolman and Sutherland 1995, Bowler et al. 2019), habitat conservation is crucial for the conservation of endangered species (Kühl et al. 2009: Hanane et al. 2016). Such knowledge allows us to intensify conservation efforts in areas where suitable habitat is disturbed and is scarce for large number of African bird species.

West Africa’s highest peak, Mount Cameroon, is one of the most important sites for globally threatened and endemic bird species in Cameroon, including Mount Cameroon Francolin (Pternistis camerunensis) and Mount Cameroon Speirops (Speirops melanocephalus). It is part of the Cameroon Mountain Endemic Bird Area (Stattersfield et al. 1998, Fotso et al. 2001), an area which continues to be degraded by natural phenomenon and anthropogenic threats. However, knowledge of the endangered species living there is limited, and no detailed work has been done to determine the effect of potential threats on their habitat preferences. Apart from the volcanic lava flows which degrade the ecosystems on Mount Cameroon, vegetation structure is also modified by the elephants (Maicher et al. 2020, Kamga et al. 2022) and the collection of the tree bark of the African Cherry (Prunus africana) for traditional medicine, which is an important source of food for some animals (Djomo Nana et al. 2014, Mayaka et al. 2021). Hunting is known to occur within this mountain, but its impact on bird populations is currently unknown (Awa II et al. 2015, Maurice et al. 2019). The combined effect of these threats can lead to serious consequences for the survival of endemic species (Titeu et al. 2016), most particular on Francolins, which are reported to be sensitive to changes in their habitats (Ssemmanda and Fuller 2005, Arlettaz et al. 2013). Therefore, we focused on one of the two bird species endemic to Mount Cameroon, the Mount Cameroon Francolin, and made efforts to provide up-to-date insight into their habitat preferences and potential threats to their population.

Based on what is known, Mount Cameroon Francolin population is undergoing a severe decline (Mcgowan et al. 2012, IUCN 2023, Tian et al. 2018). It inhabits dense undergrowth within primary forest and clearings between 850–2100 m a.s.l. on Mount Cameroon. The Francolin is a monotypic species and similar to the Noble Spurfowl (Pternistis nobilis) from the distant Ruwenzori and Kivu mountains in Uganda and the Democratic Republic of Congo (Hall 1963, Suinyuy 2006). Hall (1963) hypothesized that montane spurfowls were once widespread in lowland cold and dry forest. However, as forest decreased in extent during warmer, wetter times, the Cameroon and Congo/Ugandan forms retreated into the now much more restricted montane forest and were cut off from one another and may have been isolated for millennia. This pattern of disjunct distribution may have allowed a greater possibility for divergence (Hall 1963). Mount Cameroon Francolin are endemic to Mount Cameroon and have a very small range (ca. 200 km²), where they face increasing threats from hunting and deliberate burning of their habitat, causing the retreat of the forest/savannah boundary in some places (Bayly and Motombe 2004). The gradual disappearance of African Cherry whose seeds are believed to be eaten and disseminated by the Francolin together with excessive debarking for international trade represent another threat jeopardizing the conservation of Francolins (Stewart 2003, Mayaka et al. 2021).

Though the species is classified as endangered (BirdLife International 2023) and fully protected under Cameroon’s legislation (MINFOF 2020), little is known about its biology (Fotso et al. 2001, Sinclair and Ryan 2003, BirdLife International, 2005). Previous fieldwork described the species as common with an estimated number of 600–1700 mature individuals (Collar and Stuart, 1985). The current population status, trend, and habitat preferences are unknown. A recent study indicates that the Mount Cameroon Francolin was observed within the following types of habitats: submontane forest and montane forest, montane scrub (Mayaka et al. 2021), and forest modified by elephants (Guetse 2017) which population is increasing.

As previous findings are based on limited data, we do not have robust data on Mount Cameroon Francolin habitat preferences, its geographical distribution, and responses to human threats. Thus, the aim of our study was to combine available data (Mayaka et al. 2021) with new data on Francolin distribution, then expand field data collection along the south and south-western slopes of Mount Cameroon, test the effects of (i) vegetation structure (tree cover and loss) and chosen species abundance, (ii) presence/absence of human threats (logging of the trees or installing the hunting traps), (iii) changes of distribution among dry and wet season, and (iv) distribution along the elevational gradient.

METHODS

Study area

We conducted fieldwork in the montane and mid-elevation forest between altitudes of 800 and 2500 m a.s.l. in Mount Cameroon National Park (4.055° - 4.378° N and 9.031°-9.294° E, Fig. 1). It is situated on the coast of Southwest Cameroon, within the Gulf of Guinea. Its climate is maritime and equatorial with an average annual temperature of 25.5–27˚C (Fraser et al. 1998, Miavita, 2011). There are two seasons, a dry season from November to February and a rainy season from March to October. Average annual precipitation reaches 13,000 mm (Fraser et al. 1998). The National Park was established in 2009, for the protection of endangered fauna and flora species from uncontrolled harvesting and poaching (MINFOF 2014). Mount Cameroon represents a 58,178-ha block of intact mountain environment and is one of the most diverse ecosystems in Cameroon (MINFOF 2014). It is an active volcano with large swaths of rugged terrain difficult to access due to extensive lava flows and volcanic craters (Proctor et al. 2007). The forest remaining on the mountain is a representative of two main habitats, lowland and montane forest. At high altitudes, there are submontane and montane forest zones (Cable and Cheek, 1998, Cheek et al. 2000). The vegetation at low altitudes comprises a coastal area of mangroves at inland dense humid evergreen Guinean - Congolian forest of the Atlantic Biafran type. The mid-elevation and montane forest between 900 and 2500 m a.s.l. is recognized as a home for endemic and threatened species such as Mount Cameroon Francolin (BirdLife International 2000) and African Cherry (Foaham et al. 2009). Mount Cameroon has experienced increased mortality of trees compared to other vegetation types due to fire and volcanic activities (Forboseh et al. 2011). Hunting and collection of non-timber forest products is still observed there (Djomo Nana and Tchamadeu 2014, Mayaka et al. 2021). While conducting field work, snares, traps, and hunters were encountered and there was evidence of animal (including Francolins) mortality.

Bird sampling

In this study, we enlarged the dataset published in Mayaka et al. (2021). We increased the sample size from 86 to 240 census points and we improved the geographical coverage on the south and south-western slope of Mount Cameroon. Here, the Mayaka et al. (2021) dataset and new data are combined in the analyses. Overall, data were collected in July 2016, 2021, and 2022, and February 2022 during the period of rainy (July) and dry (February) season. We used the point count method (Bibby et al. 2000, Sedláček et al. 2015) with points located 300 m apart. We collected data between 6:30 and 11:30 a.m. and in the evening from 3:30 and 5:30 p.m., but mostly during the morning, since the birds are more active during these times (Bibby et al. 2000). We only collected data during periods without heavy rains. We combined the acoustic monitoring technique with the playback since the Mount Cameroon Francolin is an elusive species and difficult to observe in the field (Mayaka et al. 2021, Fuller et al. 2012). We then moved along an established route, stopping at specific points and counting all seen and heard birds. At each study station, we pre-listened to the natural vocal activity of the species for a period of three minutes. In the absence of a natural response, we played the song of the Francolin obtained from the Chappuis collection (Chappuis, 2000) via an audio player connected to an amplified loudspeaker for 20 s within three repetitions. Then we waited for five minutes to follow the response or responses of individuals of the species around each point.

Habitat characterization

We visually estimated vegetation cover within a 50 m radius around each counting station. We estimated the vegetation cover for each of the following vertical strata: 0–1 m, 1–3 m, 3–5 m, 5–10 m, and above 10 m. For each vegetation layer we noted a percentage of cover, representing the density of the vegetation within the stratum. This method has been successfully used to characterize the vegetation elsewhere (Reif et al. 2007). At each counting point in 2022, we also collected the following environmental variables; (i) abundance of the African Cherry around the counting point (absent, low ~ when number is ≤ 7 trees and abundant ~ when number is 7 > trees), (ii) logging intensity of African Cherry by counting number of stumps (0–7), presence/absence of Syzigium trees (0/1), (iii) presence/absence of Macaranga trees (0/1), presence/absence of Polyscias fulva trees (0/1), note that we selected the tree species based on interviews with local people who suggested that the fruits and seeds of these species are part of the Francolins’ diet, (iv) presence/absence of wildlife traps installed by humans, presence of traps increases probability of human occurrence and potentially human caused mortality to Francolins (0/1); (v) presence/absence of human activities like farmland and non-timber forest products collection by direct observation (0/1), (vi) bushfire presence/absence (0/1) and (vii) geographical coordinates (longitude, latitude, elevation). Data was collected once during each season.

Statistical analyses

We modeled Francolin occurrence using a binomial general linear model (glm), and we separately analyzed four main datasets according to the sampling period (all years, year 2022) and dependent variable (presence/absence of Francolin and Francolin group size). The first two datasets from all the years were based on dependent variable (presence/absence and group size of Francolins (0–5) within the radius around the counting points (n = 240 counting stations). These two dependent variables were chosen to simultaneously cover the whole spatial distribution of the species range and the group size. We modeled Francolin occurrence using binomial glm with a Logit link function in R 4.0.5 (R Core Team 2021) to uncover the effect of following independent variables: 1) longitude (i.e., longitudinal rainfall gradient), (2) month (July/February), 3) year (2016, 2021, 2022), 4) whether or not the point fell within the National Park (0/1), 5) main vegetation (forest, farmland, and pasture), 6) main threats: poaching, forest destruction, wild fires, agriculture (0/1), 7) tree cover (9–95%) and 8) tree loss (0–7%) within a radius 150 m around the counting point. The last two mentioned parameters were estimated using ArcGIS Pro software (ESRI 2011) using raster layer of forest loss provided at 30m resolution from https://www.globalforestwatch.org/map (Hansen et al. 2013). For the second dataset on group size, we used the same independent variables as before, but we also added the abundance of African Cherry (absent, low, and abundant). For this analysis, we used a zero inflated Poisson glm using a Log link function using the glmmTMB package in R software. We performed Spearman rank correlation test among continuous independent variables (longitude, tree cover, and tree loss) using Statistica 14 software (TIBCO Software Inc. 2020). We did not find any significant relationship among these independent variables (for all comparisons, P > 0.05). For models with categorical/binomial independent variable (abundance of African Cherry, presence/absence of threats, and wet/dry season), we calculated odds ratio based on numbers of cases within each category using the epitools package in R software (Sergeant, 2018). We generated contrast in probability of Francolin occurrence between high African Cherry abundance and compared to when it was not present.

The two further analyses from the dataset gathered from the year 2022 differed from previous analyses. We again used the two independent variables (absence/presence and group size of Francolins) with the same distribution of these dependent variables as mentioned above (n = 52 counting points). However, for both analyses we only used the following independent variables in our glm models: 1) elevation above the sea level, 2) number of African Cherry trees (0–7), 3) number of cut African Cherry trees (0–7), 4) presence/absence of Syzigium trees (0/1), 5) presence/absence of human traps (0/1), 6) presence/absence of Macaranga trees (0/1), 7) presence/absence of Polyscias fulva tree (0/1), 8) presence/absence of Elimnia trees (0/1), and 9) presence/absence of human activity (0/1). We performed Spearman rank correlation test among continuous independent variables (elevation, number of African Cherry trees, and number of cut African Cherry trees) using Statistica 14 software (TIBCO Software Inc., 2020). We found only one statistically significant positive relationship between elevation and number of African Cherry trees (r_s = 0.69, P < 0.05). For other comparisons, we did not find any statistically significant relationship (for all comparisons, P > 0.05).

For each dependent variable, we first performed multimodel comparison (Burnham et al. 2011) using dredge function (package MuMIN) in R 4.0.5 software (R Core Team 2021) to find which variables should be further tested based on AIC differences (Appendix 1–4, Tables S1–4). Firstly, we created an intercept only for null model without independent variables. Then we created an alternative model for each independent variable and we used the anova function to test the significance of each model (Chambers and Hastie 1992) in R 4.0.5 software. Graphs were prepared using Statistical 14 software. The curve fitting for the relationship between group size and tree loss values within the buffers were computed using non-linear estimation in Statistica 14 software. If necessary, we used post hoc tests for multilevel independent variables using glht (package multcomp) in R 4.0.5 software to find significance for differences among each pair of levels.

Based on the understanding of drivers of Francolin presence/absence, we estimated potential Mount Cameroon Francolin population size within the occupied elevational range 950–2350 m a.s.l. using ArcGIS Pro software. We used a DEM model (Japan Aerospace Exploration Agency 2021) to specify the extent of this elevational range on Mount Cameroon. Additionally, we employed random forest-based classification and regression techniques to predict the species occurrence within the elevational range (Breiman et al. 2017). We treated the presence/absence of Mount Cameroon Francolins as the dependent variable and tree cover (0–100%) derived from the Hansen et al. (2013) raster, and elevation (m) from the aforementioned DEM as explanatory variables. The analysis calculated a 90% prediction interval around each predicted value. Based on our results, the predicted species occurrence for tree cover was between 51 and 99%. Further criterion was elevation of points with presence/absence of the species that excluded presence of the species within certain elevational band. Based on these two criterions, regression trees were built, and we gathered the final prediction map. Within the predicted area of occurrence, we further created random points with the distance 300 m from each other to ensure the radius 150 m around each point (i.e., home range size). In total, there was possible to locate 1307 random points within predicted area. Based on our results, we further assessed the mean (± s.d.) group size (2.0 ± 1.0 inds./group). By multiplying these two values, we obtained a final population size estimate within the predicted area.

RESULTS

Factors influencing occurrence of Francolins

Based on the analyses of the whole dataset, the occurrence of Francolins was influenced by abundance of African Cherry, presence of human threats, and season (Table 1). The highest occupancy was found when African Cherry abundance was high and the lowest in its absence (Fig. 2a). Using post hoc comparisons, we found statistical significance among all pairs of the levels of African Cherry abundance (P at least 0.0001). Francolins were 2.87 times more likely to be present when African Cherry abundance was high compared to when it was absent (Appendix 5, Table S5). Simultaneously, we found that presence of human threats significantly decreased probability of Francolins’ presence (Fig. 2b). The probability of Francolin occurrence was 0.08 times lower if human threats were present (Appendix 6, Table S6). Finally, percentage of Francolin presence was higher during the dry season than the wet season (Fig. 2c). During the wet season, the probability of Francolin occurrence was 2.27 times lower compared to the dry season (Appendix 7, Table S7).

Factors influencing size of Francolin groups

Similar results were obtained when we tested the effect of environmental variables on Francolin group size. However, we also found a significant negative effect of tree loss within the buffers around the survey points on group size (Fig. 3a). Furthermore, it was again revealed that groups were bigger during the dry season than the wet season (Fig. 3b) and there was a complete absence of groups when human threats were present (Fig. 3c). Finally, results revealed that the group size continually increased with increasing abundance of African Cherry (Fig. 3d). Using post hoc comparisons we found statistically significant difference among all pairs of the levels of African Cherry abundance (P at least 0.00001).

Based on the 2022 data, we found that both Francolin occurrence and group size were affected by elevation, and group size was further affected by presence of human traps (Table 1). The absence of Francolins was mainly recorded in the lower elevations (median 1621 m a.s.l.) compared to the higher elevations with Francolin presence (median 2018 m a.s.l., Fig. 4a). This corroborates the finding that groups of one to three individuals were found between the elevations of 1800 and 2200 m a.s.l. (Fig. 4b). Using post hoc comparisons we compared groups of different sizes, we found only statistical significance between group size zero and one (P at least 0.024), differences among other pairs of group sizes were not statistically significant (P at least 0.554). During the survey, no Francolin were observed in the lowland forest (from 800 m a.s.l.) as in the past, using records from altitudes between 1100 and 2200 m a.s.l. No groups were recorded at localities where the human traps were found compared to other sites with absence of these traps (Fig. 4c).

Francolin population size estimate

Finally, we estimated Francolin density and population size within the occupied elevational range and suitable tree cover values (Appendix 8, Fig. S1). Our model predicted that Mount Cameroon Francolin occupied 1307 cells within the studied area. By multiplying number of points with predicted occurrence with mean group size, we obtained a final population size estimate of 2614±261 individuals. Based on the results of the analysis, tree cover had importance 14.0 (49.6%) and elevation had importance 14.2 (50.4%).

DISCUSSION

Our results suggest that the spatial distribution of Mount Cameroon Francolins is influenced by the occurrence of African Cherry within their occupied range of elevations, i.e., between 1800 and 2200 m a.s.l. We provide new evidence that the geographical distribution of Francolins is strongly affected by the presence of human threats, and this varied between dry and wet season. Finally, based on the newly collected data we estimated a population size of the species to be 2614 individuals in total.

Our data confirm previous observations of Mayaka et al. 2021, that African Cherry influences the occurrence of Francolins. It also agrees with local Indigenous knowledge from recent interviews of people regularly visiting the forest (Guetse et al., unpublished data) that the Mount Cameroon Francolin consumes the fruits of African Cherry. This is not surprising, because previous studies indicated that fruits of African Cherry are eaten by a wide range of animals endemic to the highlands (Maisels and Forboseh 1999, Stewart 2003). The seeds are most likely dispersed by birds and primates (Farwig et al. 2006). Furthermore, within the area of Mount Cameroon with absence of African Cherry (such as Njonji area), no Francolin have been observed despite previous bird surveys (Mayaka et al. 2021). This result highlights the ecological link between African Cherry and the Francolin or suggests same ecological requirements of the species. Generally, the distribution and abundance of the birds depends on foraging preferences for tree species that provide habitat requirements determined by morphological and behavioral adaptations of each bird species for food procurement and differences in their prey abundance among the tree species (e.g., Gabbe et al. 2002, Tamungang et al. 2016a). We suspect that the Mount Cameroon Francolins disseminate their seeds (Mayaka et al. 2021). However, further studies are needed to determine in greater detail the ecological relationships between African Cherry and Francolins. On Mount Cameroon, the Mount Cameroon Francolins are found mostly at the same altitudinal range (850–2100 m a.s.l.) as the African Cherry (900–2500 m a.s.l.; Foaham et al. 2009). African Cherry is considered vulnerable but widespread Afromontane hardwood species that is traded internationally for medicinal use of their barks (Cunningham and Mbenkum 1993, Vinceti et al. 2013). Our findings highlight the need for sustainable management of Cherry harvest to support the Francolin population.

Using the available data, we detected Francolins between 950 and 2350 m a.s.l., with higher frequency of records from the montane forest within the belt of elevations between 1800 and 2200 m. Such findings corroborate with the previous observations (BirdLife International 2000, Mayaka et al. 2021). Based on our estimate within the range of elevations (950–2350 m a.s.l.), there are approximately 2614 individuals of the species. This is a conservative estimate, but a likely relatively precise estimate of the entire population of Mount Cameroon Francolin, it is different from previous estimates to 600–1700 mature individuals (Collar and Stuart 1985) and 700–1500 (BirdLife International 2023). In fact, our estimate is above the upper limit of previous estimates. Such a finding is promising as it suggests that the Mount Cameroon forest may still host a reasonable population of Francolins. However, more robust estimates such as mark-recapture and models that estimate detection probability are needed to confirm our estimate. Places within the montane forest occupied by the Mount Cameroon Francolin are characterized by dense undergrowth vegetation within forest patches containing large trees (more than 20 m in height). These are often cloudy and have a discontinuous canopy cover (Thomas and Cheek 1992). Theoretically, the presence of Francolins is conditioned both by the availability of their food resources and the ability to escape predators (Hanane et al. 2016). Unfortunately, detailed field knowledge of diet preferences and predators of Mount Cameroon Francolins are still missing. Neither this study, nor previous studies reported the observation of Francolins within the lowland forest (i.e., below 800 m a.s.l.). Therefore, preferences of Francolins for higher altitudes could have serious consequences on the future survival of the species. Indeed, other mountains regions of the world are already affected by climate change (Pounds et al. 1999, Engler et al. 2011) and Mount Cameroon may not be spared by environmental changes. It is also possible that the presence of Francolin is restricted to areas with presence of African Cherry, which based on our dataset occurs above 1700 m a.s.l. Unfortunately, our analyses could not distinguish between the effect of presence of African Cherry and elevation. This issue could be addressed in the future by placing transects at the same elevation in areas both with and without African Cherry. Climate change may affect the current population and distribution of the Mount Cameroon Francolin in the future, as it has been shown in other species of birds including Galliforms (Sala et al. 2000, Thomas et al. 2005, Selås et al. 2011, Urban 2015). The survival of the species will therefore depend on its ability to adapt to new climatic conditions (cf. Parmesan et al. 1999, Walther et al. 2002). Climate change vulnerability of Francolins is reinforced by its high-altitude traits, such a thermal tolerance, scarcity, habitat specialization, and ground nesting behavior (Hof et al. 2017, Hof and Allen 2018). Similar studies have shown that Mount Cameroon Francolins also have relatively low reproductive capacity (Fuller et al. 2000, Languy et al. 2019), increasing their vulnerability.

Our results also confirmed that the presence of the Francolins is negatively affected by human threats. We found a significant negative effect of tree loss on group size of Francolins. The Mount Cameroon Francolin were recorded only in natural habitat patches, and they have never been observed in agroforest plantations or in farmland (BirdLife International 2000, Mayaka et al. 2021). This suggests that habitat loss may have reduced its population size and suggests that Francolins depend on pristine forest which are highly threatened. The logging and regular burning of the grassland around the montane forest potentially leads to loss of the Francolins’ habitat (Forboseh et al. 2011, MINFOF 2014, Guetse 2017). Consequently, Francolins have disappeared in portions of Mount Cameroon, exacerbated by the disappearance of seeds of African Cherry and their excessive debarking for international trade (Stewart, 2003). Before the creation of the Mount Cameroon National Park, the forest was degraded by fires and agricultural encroachment, uncontrolled hunting, and African Cherry exploitation (Djomo Nana and Tchemadeu, 2014).

Our study also revealed that no Francolins were recorded at localities where human traps were found compared to sites with absence of these traps. The traps being used are snares. Francolins, along with other animals, are being caught by these snares because they are non-selective. Other studies have observed that hunting effects are generally more detrimental outside of protected areas (Benitez-Lopez et al. 2017), but the fact remains that overhunting is ubiquitous within protected areas in Amazonia, Asia, and Africa (Benitez-Lepez et al. 2017). Several authors have affirmed that hunting is the major cause of the decline of larger undergrowth birds such as the Mount Cameroon Francolin (Bayly and Motombe 2004, Awa II et al. 2015, Maurice et al. 2019). Our results corroborate with similar results obtained during a study on the Black Francolin (Francolinus Francolinus), which revealed that illegal and excessive hunting contributes to the reduction of the number and the range of this species (Nastaran et al. 2009). Several conservation measures have been proposed for Mount Cameroon Francolins, including the reduction of hunting and trapping as well as the total cessation of bush fires on grassland around montane forest (BirdLife International 2000), yet none of these measures have been sufficiently implemented so far.

Interestingly, we found apparent seasonal variation in the numbers of Francolins between the wet and dry seasons. The increased abundance of Francolins during the dry season can be explained by the increased availability of essential resources during this period, such as insects, tree seeds and suitable nesting sites. Indeed, in most of tropical rainforests, the availability of primary food resources, including new leaves, flowers, seeds, and nest sites fluctuates seasonally (Van Schaik et al. 1993). There is evidence of seasonal variation in abundance and changes in bird habitat at several tropical sites (Williams and Middleton 2008). However, the situation of the Mount Cameroon Francolin is a bit specific. Seasonal changes in diversity and abundances typical, for instance, in savanna species (Bourliere et al. 1983), result from inter-seasonal movements of individuals (Bronikowski et al. 1996, Alberts et al. 2005). The information about movements of Francolins is not available, but we do not expect that they are moving across large spaces such as between lower and higher parts of the mountain. We hypothesize that they are not leaving their mountain range at a certain period of the year and are likely non-migratory species. Thus, the observed differences between dry and wet season might be because of differences in activity patterns of the birds or small-scale habitat preferences rather than changes in their geographical movements. Indeed, according to Toledo-Lima et al. (2014), many species are more detectable during the breeding period which in many cases coincides with the dry season. In Cameroon, most Francolin species breed at the beginning of the dry season (Tamungang, personal observation). This is the season when most villagers go out to collect nestlings and eggs of the bird species in forest stands and fallowed farms around villages. This likely holds also for Mount Cameroon Francolin with breeding period between October and December (McGowan et al. 2020). Nevertheless, future research that accounts for species detectability (Bailey et al. 2014) is essential. Our results also agree with observations obtained during a study on the vertical stratification of birds within the Mount Cameroon in which Chmel et al. (2021) ascertain the seasonal influences on bird abundance and species richness.

In conclusion, we provided new data on the spatial and seasonal distribution of threatened Mount Cameroon Francolin and their current population size. Our analyses revealed that they respond to human habitat modification and activities. Interestingly, we confirmed a strong positive link between African Cherry and Francolin occupancy and concluded that the bird likely eats fruits and is a potentially important seed disperser for this tree. Our study also highlights the lack of important ecological information on the species, such as breeding data, diet composition, or assessment of potential predators, which are crucial for understanding life of this rare and threatened species. Our data suggest that the small population size, limited geographic extent, and response to human activities suggest the need for careful management of its habitat that should be focused mostly on elimination of adverse human activities such as (i) agricultural land expansion, (ii) poaching, and (iii) human induces wildfires.

LITERATURE CITED

Alberts, S., J. Hollister-Smith, R. Mututua, S. Sayialel, P. Muruthi, J. Warutere, and J. Altmann. 2005. Seasonality and long-term change in a savanna environment. Pages 157-196 in D. Brockman and C. Schaik, editors. Seasonality in primates: studies of living and extinct human and non-human primates. Cambridge University Press, Cambridge, UK. https://doi.org/10.1017/CBO9780511542343.007

Arlettaz, R., P. Patthey, and V. Braunisch. 2013. Impacts of outdoor winter recreation on alpine wildlife and mitigation approaches: a case study of the black grouse. Pages 137-154 in C. Rixen and A. Rolando, editors. The impacts of skiing and related winter recreational activities on mountain environments. Bentham Science Publishers. https://doi.org/10.2174/9781608054886113010011

Awa II, T., S. A. Tamungang, E. M. Ashute, and L. C. Djieto. 2015. Preliminaries on population status and distribution of the endangered Mount Cameroon Francolin, Pternistis camerunensis. Gallinformed Newsletter 10:13-17.

Bailey, L. L., D. I. MacKenzie, and J. D. Nichols. 2014. Advances and applications of occupancy models. Methods in Ecology and Evolution 5(12):1269-1279. https://doi.org/10.1111/2041-210X.12100

Benıtez-Lopez, A., R. Alkemade, A. M. Schipper, D. J. Ingram, P. A. Verweij, J. A. J. Eikelboom, and M. A. J. Huijbregts. 2017. The impact of hunting on tropical mammal and bird populations. Science 356 (6334):180-183. https://doi.org/10.1126/science.aaj1891

Beresford, A. E., F. J. Sanderson, P. F. Donald, I. J. Burfield, A. Butler, J. A. Vickery, and G. M. Buchanan. 2019. Phenology and climate change in Africa and the decline of Afro‐Palearctic migratory bird populations. Remote Sensing in Ecology and Conservation 5(1):55-69. https://doi.org/10.1002/rse2.89

Bibby, C. J., N. D. Burgess, D. A. Hill, and S. Mustoe. 2000. Bird Census Techniques. Elsevier.

BirdLife International 2000. Threatened Birds of the World. Lynx Editions and BirdLife International, Barcelona, and Cambridge, UK.

BirdLife International 2013. Multiple threats are driving threatened birds towards extinction in Africa. https://www.birdlife.org

BirdLife International 2023. Species factsheet: Pternistis camerunensis. https://datazone.birdlife.org/species/factsheet/mount-cameroon-spurfowl-pternistis-camerunensis

Bayly, N., and F. N. Motombe. 2004. Final Report of an Ornithological Survey of Mount Cameroon. Wildlife Conservation Society.

Bowler, D. E., H. Heldbjerg, A. D. Fox, M. de Jong, and K. Böhning‐Gaese. 2019. Long‐term declines of European insectivorous bird populations and potential causes. Conservation Biology 33(5):1120-1130. https://doi.org/10.1111/cobi.13307

Bourliere, F., and M. Hadley. 1983. Present-day savannas: an overview. Pages 1-17 in F. Bourliere, editor. Ecosystems of the world 13: tropical savannas. Elsevier Scientific, Amsterdam.

Breiman, L., J. H. Friedman, R. A. Olshen, and C. J. Stone 2017. Classification and regression trees. Routledge, New York, New York, USA. https://doi.org/10.1201/9781315139470

Bronikowski, A., and J. Altmann 1996. Foraging in a variable environment: weather patterns and the behavioral ecology of baboons. Behavioral Ecology and Sociobiology 39:11-25. https://doi.org/10.1007/s002650050262

Burnham, K. P., D. R. Anderson, and K. P. Huyvaert. 2011. AIC model selection and multimodel inference in behavioral ecology: some background, observations, and comparisons. Behavioral Ecology and Sociobiology 65:23-35. https://doi.org/10.1007/s00265-010-1029-6

Cable, S., and M. Cheek. 1998. The plants of Mount Cameroon: A conservation checklist. Royal Botanic Gardens, Kew, UK.

Chambers, J. M., and T. J. Hastie. 1992. Statistical Models in S, Wadsworth and Brooks/Cole.

Chappuis, C. 2000. Oiseaux d’Afrique (African Bird Sounds) 2. West and Central Africa. 11 CDs. Paris, France: Societé d’Etudes Ornithologiques de France.

Cheek, M., J. M. Onana, and B. J. Pollard. 2000. The plants of the Mount Oku and the Ijim Ridge: a conservation checklist.

Chmel, K., M. S. Kamga, T. Awa II, F. Luma, G. Uceda-Gomez, D. Hořák, J. Mlíkovský, L. L. Molua, J. Riegert, and Š. Janeček, 2021. Vertical stratification and seasonal changes of the avian community in Mount Cameroon lowland rainforest. African Journal of Ecology 59(3):655-666. https://doi.org/10.1111/aje.12877

Collar, N. J., and S. N. Stuart 1985. Threatened birds of Africa and related islands: the ICBP/IUCN Red Data Book. Cambridge, UK: International Council for Bird Preservation, and International Union for Conservation of Nature and Natural Resources.

Cooper, T. J., A. M. Wannenburgh, and M. I. Cherry. 2017. Atlas data indicate forest dependent bird species declines in South Africa. Bird Conservation International 27(3):337-354. https://doi.org/10.1017/S095927091600040X

Cunningham, A. B., and F. T. Mbenkum. 1993. Sustainability of harvesting Prunus africana in Cameroon: a medicinal plant in international trade. People and Plants working paper 2. UNESCO, Paris, France.

De Klerk, H. M., T. M. Crowe, J. Fjeldså, and N. D. Burgess. 2002. Biogeographical patterns of endemic terrestrial Afrotropical birds. Diversity and Distributions 8(3):147-162. https://doi.org/10.1046/j.1472-4642.2002.00142.x

Djomo Nana, E., and N. N. Tchamadeu. 2014. Socio-economic impacts of protected areas on people living close to the Mount Cameroon National Park. https://doi.org/10.2305/IUCN.CH.2014.PARKS-20-2.EDN.en

Djomo Nana E., O. Sedláček, N. Bayly, T. Albrecht, M. Ferenc, J. Reif, F. Motombi, F. and D. Hořák. 2014. Comparison of avian assemblage structures in two upper montane forests of the Cameroon volcanic line: lessons for bird conservation. Biodiversity and Conservation 23:1469-1484. https://doi.org/10.1007/s10531-014-0677-7

Dolman, P. M., and W. J. Sutherland. 1995. The response of bird populations to habitat loss. Ibis 137:S38-S46. https://doi.org/10.1111/j.1474-919X.1995.tb08456.x

Engler, R., C. F. Randin, W. Thuiller, S. Dullinger, N. E. Zimmermann, M. B. Araújo, P. B. Pearman, G. Le Lay, C. Piedallu, C. H. Albert, P. Choler, G. Coldea, X. De Lamo, T. Dirnböck, J. C. Gégout, D. Gómez-García, J. A. Grytnes, E. Heegaard, F. Høistad, D. Nogués-Bravo, S. Normand, M. Puşcaş, M. T. Sebastià, A. Stanisci, J. P. Theurillat, M. R. Trivedi, P. Vittoz. and A. Guisan. 2011. 21st century climate change threatens mountain flora unequally across Europe. Global Change Biology 17:2330-2341. https://doi.org/10.1111/j.1365-2486.2010.02393.x

ESRI. 2011. ArcGIS desktop: release 10. Environmental Systems Research Institute, Redlands, California, USA.

Farwig, N., K. Böhning-Gaese, and B. Bleher. 2006. Enhanced seed dispersal of Prunus africana in fragmented and disturbed forests? Oecologia 147:238-252 https://doi.org/10.1007/s00442-005-0288-9

Fjeldsaå, J., D. Ehrlich, E. Lambin, and E. Prins. 1997. Are biodiversity ‘hotspots’ correlated with current ecoclimatic stability? A pilot study using the NOAA-AVHRR remote sensing data. Biodiversity and Conservation 6:401-422. https://doi.org/10.1023/A:1018364708207

Foaham, B., S. Dagobert, V. Ingram, and A. Awono. 2009. Inventaire de prunus africana dans les provinces du Sud-ouest et du Nord-ouest du Cameroun, Novembre 2007-Novembre 2008. CIFOR.

Forboseh, P. F., T. C. H. Sunderland, J. A. Comiskey, and M. Balinga. 2011. Tree population dynamics of three altitudinal vegetation communities on Mount Cameroon (1989-2004). Journal of Mountain Science 8:495-504. https://doi.org/10.1007/s11629-011-2031-9

Fotso, R., F. Dowsett-Lemaire, R. J. Dowsett, P. Scholte, M. Languy, and C. Bowden. 2001. Pages 133-150 in L. D. C. Fishpool and M. I. Evans, editors. Important Bird Areas.

Fraser, P. J., J. B. Hall, and J. R. Healey. 1998. Climate of the Mount Cameroon Region: Long and medium term rainfall temperature and sunshine data. School of Agricultural and Forest Sciences Publication N˚16. University of Wales, Bangor, UK.

Fuller, R. A., J. P. Carroll, and P. J. K. McGowan. 2000. Partridges, quails, Francolins, snowcocks, guineafowl, and turkeys. Status survey and conservation action plan 2000-2004. IUCN and World Pheasant Association, Gland, Switzerland and Cambridge, UK.

Fuller, R. A., P. Akite, J. B. Amuno, C. L. Fuller, J. M. Ofwono, G. Proaktor, and R. Ssemmanda. 2012. Using playback of vocalisations to survey the Nahan’s Francolin, a threatened African forest galliform. Ostrich 83:1-6. https://doi.org/10.2989/00306525.2012.677510

Gabbe, A. P., S. K. Robinson, and J. D. Brawn. 2002. Tree-species preferences of foraging insectivorous birds: implications for floodplain forest restoration. Conservation Biology 16(2):462-470. https://doi.org/10.1046/j.1523-1739.2002.00460.x

Gaston, K. J., T. M. Blackburn, and K. K. Goldewijk. 2003. Habitat conversion and global avian biodiversity loss. Proceedings of the Royal Society of London. Series B: Biological Sciences 270(1521):1293-1300. https://doi.org/10.1098/rspb.2002.2303

Guetse, F. 2017. Modification du couvert végétale entre 2008 et 2015 et effet des caractéristiques de l’habitat sur l’occurrence du Francolin du mont Cameroun. Mémoire de master, Université de Dschang.

Hall, B. P. 1963. The Francolins, a study in speciation. Bulletin of the British Museum (Natural History) 10:8-24. https://doi.org/10.5962/bhl.part.20525

Hanane, S., N. Alahyane, M-A. El. Magri, and A. Qninba. 2016. Habitat preference of the sole wild population of Francolinus bicalcaratus ayesha in the palearctic: implications for conservation and management. Revue d’Ecologie (Terre et Vie) 71(3):288-297. https://doi.org/10.3406/revec.2016.1852

Hansen, M. C., P. V. Potapov, R. Moore, M. Hancher, S. A. Turubanova, A. Tyukavina, D. Thau, S. V. Stehman, S. J. Goetz, T. R. Loveland, A. Kommareddy, A. Egorov, L. Chini, C. O. Justice, and J. R. G. Townshend. 2013. High-resolution global maps of 21st-century forest cover change. Science 342:850-53. https://doi.org/10.1126/science.1244693

Hof, A. R., G. Rodríguez‐Castañeda, A. M. Allen, R. Jansson, and C. Nilsson. 2017. Vulnerability of Subarctic and Arctic breeding birds. Ecological Applications 27(1):219-234. https://doi.org/10.1002/eap.1434

Hof, A. R., and A. M. Allen. 2019. An uncertain future for the endemic Galliformes of the Caucasus. Science of the Total Environment 651(2019):725-735. https://doi.org/10.1016/j.scitotenv.2018.09.227

Hořák, D., V. R. Clark, K. Y. Njabo, and J. Fjeldså. 2023. Biodiversity across Afromontane environments. Frontiers in Ecology and Evolution 10:1080119. https://doi.org/10.3389/fevo.2022.1080119

IUCN. 2023. The IUCN Red List of Threatened Species. https://www.iucnredlist.org

Japan Aerospace Exploration Agency. 2021. ALOS World 3D 30 meter DEM. V3.2, Jan 2021. Distributed by OpenTopography.

Kamga, S. M., S. A. Tamungang, T. Awa, F. L. Ewome, F. N. Motombi, D. Hořák, and J. Riegert. 2022. The importance of forest elephants for vegetation structure modification and its influence on the bird community of a mid-elevation forest on Mount Cameroon, West-Central Africa. Diversity 14:227. https://doi.org/10.3390/d14030227

Kühl, H., F. Maisels, M. Ancrenaz, and E. Williamson. 2009. Lignes directrices pour de meilleures pratiques en matière d’inventaire et de suivi des populations de grands singes. Groupe de spécialiste des primates de la CSE de l’Union Mondiale pour la Nature, Gland, Suisse.

Languy, M. 2019. The birds of Cameroon. Their status and distribution. Series studies in afrotropical zoology 299. Royal Museum for Central Africa, Tervuren.

Maicher, V., S. Delabye, M. Murkwe, J. Doležal, J. Altman, I. N. Kobe, and R. Tropek. 2020. Effects of disturbances by forest elephants on diversity of trees and insects in tropical rainforests on Mount Cameroon. Scientific Reports 10(1):21618. https://doi.org/10.1038/s41598-020-78659-7

Maisels, F., and P. Forboseh. 1999. Phenology of the major tree and shrub species of the Kilum-Ijim forest. Ecological monitoring programme. Project, K.-I.F. Cameroon, BirdLife International and the Ministry of Environment and Forestry.

Maurice, M. E., N. A. Fuashi, A. F. Zeh, E. L. Ebong, O. A. Olle, H. Konjieh, and N. F. Daizy. 2019. An exploration on the perception of the inhabitants of Muyuka municipality towards the conservation of Francolins, Southwest Region, Cameroon. LS - An International Journal of Life Sciences. https://doi.org/10.5958/2319-1198.2019.00003.4

Mayaka, T. B., F. Guetse, A. Chemura, R. Tamungang, and D. Hořák. 2021. A wake-up call: playback attraction data link numbers of Mount Cameroon Francolin Pternistis camerunensis to abundance and debarking of red stinkwood Prunus africana. Bird Conservation International 32(1):43-52. https://doi.org/10.1017/S0959270921000058

McGowan, P., L. Owens, and M. Grainger. 2012. Galliformes science and species extinctions: what we know and what we need to know. Animal Biodiversity and Conservation 35(2):321-331. https://doi.org/10.32800/abc.2012.35.0321

McGowan, P. J. K., E. de la Juana, G. M. Kirwan, and P. F. D. Boesman. 2020. Mount Cameroon Spurfowl (Pternistis camerunensis), version 1.0. In J. del Hoyo, A. Elliott, J. Sargatal, d. A. Christie, and E. de la Juana, editors. Birds of the world. Cornell Lab of Ornithology, Ithaca, New York, USA. https://doi.org/10.2173/bow.camfra2.01

Miavita. 2011. Report on Mount Cameroon Socio-economic vulnerability and resilience. Collaborative project - FP7-ENV-2007. Manufacturing. Technometrics 34(1):1-14.

MINFOF 2014. The management plan of the Mount Cameroon National Park and its peripheral zone 2015-2019. Ministry of Forest and Wildlife, Yaoundé. Ministry of Environment and Forestry 1994. A Compendium of official instruments on forest and wildlife management in Cameroon. MINEF, Yaoundé, Cameroon.

Ministère des Forêts et de Faune. 2020. Arrêté n° 0053/MINFOF du 01 Avril 2020 fixant les modalités de répartition des espèces animales en classe de protection.

Nastaran, H., A. Tayebeh, and N. Gholamreza, and S. 2009. Distribution, population, and ecology of Black Francolin Francolinus Francolinus bogdanovion the Sistan Plain, in relation to plant coverage and drought. Podoces 4(1):28-36.

Parmesan, C., N. Ryrholm, C. Stefanescu, J. K. Hill, C. D. Thomas, H. Descimon, B. Huntley, L. Kaila, J. Kullberg, T. Tammaru, W. J. Tennent, J. A. Thomas, and M. Warren. 1999. Poleward shifts in geographical ranges of butterfly species associated with regional warming. Nature 399:579-583. https://doi.org/10.1038/21181

Pounds, J. A., M. P. L. Fogden, and J. H. Campbell. 1999. Biological response to climate change on a tropical mountain. Nature 398:611-615. https://doi.org/10.1038/19297

Proctor, J. and I. D. Edwards. 2007. Zonation of forest vegetation and soils of Mount Cameroon. West Africa. Plant Ecology 192:251-269. https://doi.org/10.1007/s11258-007-9326-5

R Core Team. 2021. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/

Reif, J., O. Sedláček, D. Hořák, J. Riegert, M. Pešata, Z. Hrázský, and Š. Janeček. 2007. Habitat preferences of birds in a montane forest mosaic in the Bamenda Highlands, Cameroon. Ostrich 78:31-36. https://doi.org/10.2989/OSTRICH.2007.78.1.5.49

Ricklefs, R. E. 2013. Habitat‐independent spatial structure in populations of some forest birds in eastern North America. Journal of Animal Ecology 82(1):145-154. https://doi.org/10.1111/j.1365-2656.2012.02024.x

Riegert, J., K. Chmel, J. Vlček, Z. Hrázský, O. Sedláček, S. Grill, J. Reif, and D. Hořák. 2021. Alarming declines in bird abundance in an Afromontane global biodiversity hotspot. Biodiversity and Conservation 30:3385-3408. https://doi.org/10.1007/s10531-021-02252-1

Sala, O. E., F. I. I. I. Chapin, J. J. Armesto, E. Berlow, J. Bloomfield, R. Dirzo, E. Huber-Sanwald, L. F. Huenneke, R. B. Jackson, A. Kinzig, R. Leemans, D. M. Lodge, H. A. Mooney, M. Oesterheld, N. L. Poff, M. T. Sykes, B. H. Walker, M. Walker, and D. H. Wall. 2000. Global biodiversity scenarios for the year 2100. Science 287:1770-1774. https://doi.org/10.1126/science.287.5459.1770

Sedláček, O., J. Vokurková, M. Ferenc, E. Djomo Nana, T. Albrecht, and D. Hořák. 2015. A comparison of point counts with a new acoustic sampling method: a case study of a bird community from the montane forests of Mount Cameroon. Ostrich 86:213-220. https://doi.org/10.2989/00306525.2015.1049669

Selås, V., G. A. Sonerud, E. Framstad, J. A. Kålås, S., Kobro, H. B. Pedersen, T. K. Spidso, and O. Wiig. 2011. Climate change in Norway: warm summers limit grouse reproduction. Population Ecology 53:361-71. https://doi.org/10.1007/s10144-010-0255-0

Sergeant, E. S. G. 2018. Epitools Epidemiological Calculators. Ausvet. http://epitools.ausvet.com.au

Sinclair, I., and P. Ryan. 2003. Birds of Africa south of the Sahara. Struik Publishers, Cape Town, South Africa.

Spooner, F. E., R. G. Pearson and, R. Freeman. 2018. Rapid warming is associated with population decline among terrestrial birds and mammals globally. Global Change Biology 24(10):4521-4531. https://doi.org/10.1111/gcb.14361

Ssemmanda, R., and R. Fuller. 2005. Assessing the status of Handsome Francolin Francolinus nobilis in Bwindi Impenetrable National Park, western Uganda. Scopus 25:41-50.

Stattersfield, A. J., N. J. Crosby, A. G. Long, and D. C. Wege. 1998. Endemic bird areas of the world. Priorities for bird conservation. BirdLife International, Cambridge, UK.

Stewart, K. M. 2003. The African cherry (Prunus africana): can lessons be learned from an over-exploited medicinal tree? Journal of Ethnopharmacology 89: 3-13. https://doi.org/10.1016/j.jep.2003.08.002

Suinyuy, T. 2006. The taxonomic status and phylogenetic relationships of some phasianoid gallinaceous gamebirds from Cameroon. Thesis, University of Cape Town, South Africa.

Tamungang, S.A., M. A. Onabid, T. Awa, and V. S. Balinga. 2016. Habitat preferences of the grey parrot in heterogeneous vegetation landscapes and their conservation implications. International Journal of Biodiversity 4:1-10. https://doi.org/10.1155/2016/7287563

Thiollay, J. M. 2006. Severe decline of large birds in the Northern Sahel of West Africa: a long-term assessment. Bird Conservation International 16(4):353-365. https://doi.org/10.1017/S0959270906000487

Thomas, D. and M. Cheek. 1992. Vegetation and Plant Species on the South Side of Mount Cameroon in the Proposed Etinde Reserve, Consultancy Report, GoC/ODA/RBG Kew.

Tian, S., J. Xu, J. Li, Z. Zhang, and Y. Wang. 2018. Research advances of Galliformes since 1990 and future prospects. Avian Research 9:1-13. https://doi.org/10.1186/s40657-018-0124-7

TIBCO Software, Inc. 2020. Data Science Workbench, version 14. https://tibco.com

Titeu, N., K. Henle, J. B. Mihoub, A. Regos, I. Geijzendorffe, W. Cramer, P. Verburg, and L. Brotons. 2016. Biodiversity scenarios neglect future land‐use changes. Global Change Biology 22(7):2505-2515. https://doi.org/10.1111/gcb.13272

Toledo-Lima, G., P. R. B. P. Macario, O. S. Gabriel, W. R. Sugliano, J. Telino, and A. F. Luiz. 2014. Richness, composition and trophic groups of an avian community in the Pernambuco Endemism Centre, Alagoas, Brazil. Anais da Academia Brasileira de Ciências Printed version ISSN 0001-3765 / Online version ISSN 1678-2690. https://doi.org/10.1590/0001-3765201420130129

Van Schaik, C. P., J. Terborgh, and S. J. Wright. 1993. The phenology of tropical forests: adaptive significance and consequences for primary consumers. Annual Review of Ecology and Systematics 24(1):353-377. https://doi.org/10.1146/annurev.es.24.110193.002033

Urban, M. C., 2015. Climate change. Accelerating extinction risk from climate change. Science 348:571-573. https://doi.org/10.1126/science.aaa4984

Vinceti, B., J. Loo, H. Gaisberger, M. J. van Zonneveld, S. Schueler, H. Konrad, A. C. Kadu, and T. Geburek. 2013. Conservation priorities for Prunus africana defined with the aid of spatial analysis of genetic data and climatic variables. PLoS ONE 8(3):e59987. https://doi.org/10.1371/journal.pone.0059987

Walther, G. R., E. Post, P. Convey, A. Menzel, C. Parmesan, T. J. C. Beebee, J. M. Fromentin, O. Hoegh-Guldberg, and F. Bairlein. 2002. Ecological responses to recent climate change. Nature 416:389-395. https://doi.org/10.1038/416389a

Williams, S. E., and J. Middleton. 2008. Climatic seasonality, resource bottlenecks, and abundance of rainforest birds: implications for global climate change. Diversity and Distributions 14:69-77. https://doi.org/10.1111/j.1472-4642.2007.00418.x

Whytock, R. C., B. J. Morgan, T. Awa II, Z. Bekokon, E. A. Abwe, R. Buij, M. Virani, J. A. Vickery, and N. Bunnefeld. 2018. Quantifying the scale and socioeconomic drivers of bird hunting in Central African forest communities. Biological Conservation 218:18-25. https://doi.org/10.1016/j.biocon.2017.11.034