Diversity and altitudinal distribution of bats (Mammalia: Chiroptera) on Mount Cameroon

Abstract

Altitudinal distribution and composition of biodiversity are a major current focus in ecology and biogeography, as they provide valuable insights into how biotic communities respond to changing ecological and climatic conditions. In this study, we document bat species richness and abundance along an elevational gradient on Mount Cameroon from sea level to 2,400 m a.s.l. Bats were mist netted in four elevational ranges corresponding to four montane vegetation types: disturbed lowland forest (0–800 m a.s.l.), disturbed sub-montane forest (801–1,600 m a.s.l.), montane forest (1,601–1,800 m a.s.l.), montane scrub (1,801–2,400 m a.s.l.), during the dry and the rainy season from November 2016 to July 2018. Forty-nine days of fieldwork resulted in the capture of 566 bats belonging to six families, 17 genera and 20 species. Species richness peaked at 475 m a.s.l. and decreased at higher elevations. The vast majority of bats captured were frugivorous bats (93.99%). We also observed a difference in species that characterize lower elevations, disturbed habitats (<1,600 m a.s.l.) (including Eidolon helvum, Epomops franqueti, Megaloglossus woermanni, Micropteropus pusillus, Nanonycteris veldkampii) and those that characterize higher elevations with primary vegetation (1,601–2,400 m a.s.l.) (including Lissonycteris angolensis and Rousettus aegyptiacus). Our data indicate that some species are much more likely to be affected by ongoing anthropogenic disturbances than others because of their spatial limitation and restrictions in ecological preferences. Our data also contributes to a better description of the bat fauna of Mount Cameroon including how species are distributed at higher altitude and different montane vegetation types.

Keywords:

• Mount Cameroon
• elevational gradient
• altitudinal distribution
• species richness
• vegetation type

Introduction

Tropical montane forests (TMFs) are recognized globally as biodiversity ‘hotspots’ and a refuge of many endemic and endangered biota (Bubb et al. 2004; Richter et al. 2009; Boehmer 2011). They provide important ecosystem services, such as air quality regulation, carbon sequestration, regulation of water balance, and erosion regulation (Boehmer 2011). However, in recent decades these forests are under threat, principally from human induced deforestation (Hamilton 1995). A major consequence of deforestation and habitat degradation is the loss of biodiversity (Parrotta et al. 2012). Indeed, the widespread fragmentation of forest is known to greatly alter the diversity and abundance of bats (Cosson et al. 1999). Bats are excellent bioindicators that show measurable responses to environmental stressors such as climate change and habitat degradation which reflect the health of the entire ecosystem (Jones et al. 2009). According to Thomas et al. (2004), knowledge on altitudinal pattern of bat species richness can be used to predict some of the possible outcomes of climate change.

Indeed, mountain gradients provide ideal situations for monitoring some of the possible impacts of anthropogenic induced global climate change on biotic communities. This is because they provide variability of ecological and climatic conditions over a relatively short horizontal distance (Ashton et al. 2011). These climatic conditions create a variety of ecosystems that support different organisms generating altitudinal patterns (Hodkinson 2005). Meta-analyses compiled data on altitudinal diversity of various vertebrate taxa (McCain 2005, 2007a, 2007b, 2009, 2010; Guo et al. 2013), have recognized four distinct patterns whose predominance depends on the taxa. McCain and Grytnes (2010) detailed these patterns to include a decrease in species richness with increase in elevation, a peak in richness at mid-elevation, low plateau with a mid-peak in richness and a low plateau pattern.

Two elevational patterns of bat species richness have been acknowledged in response to altitudinal gradient depending on the climatic conditions at the mountain base: decline in richness with elevation occurs in mountains with wet, warm bases and a mid-elevational peak in mountains with arid bases (McCain 2007b). A number of interacting factors are responsible for the influence of altitudinal gradient on species richness, among which climate, plant productivity and habitat heterogeneity are most frequently cited explanations for both a linear and a humped relationships between species richness and elevation (McCain and Grytnes 2010).

Mount Cameroon is reputed for its high species richness and endemism across many taxa, a consequence of its broad elevational gradient characterized by a variety of habitat types that vary with increasing altitude. According to Maley (2002), the high level of endemism can be explained by the fact that the region was probably part of an important pleiostocene refuge. Moreover, Mount Cameroon is the only remaining area in West Africa with undisturbed natural altitudinal vegetation extending from sea level at the coast to alpine savannah at the summit (Acworth et al. 1996; Forboseh et al. 2011), and it is also a biodiversity hotspot (IUCN 2014).

Despite the mountain’s importance in conservation, its bat fauna remains poorly sampled. Previous bat surveys in the Mount Cameroon region were compiled in checklist and with one exception Fedden and Macleod (1986), surveys date back more than 40 years. Given the major recent research interest in bat fauna of Cameroon (e.g. Bakwo Fils 2009, 2010; Bakwo Fils et al. 2012; Bakwo Fils 2014; Bakwo Fils et al. 2014; Hassanin 2014; Lebreton et al. 2014; Bakwo Fils et al. 2018), as well as the ongoing threat to biodiversity in the area, it is imperative that more recent studies that document bat diversity and distribution be carried out in conservation priority area such as Mount Cameroon. Furthermore, studies in this area can help conservationists understand how bats use high altitude areas such as montane forest and scrub.

In this paper, we characterize bat species richness, abundance and distribution along an altitudinal gradient correlated with four montane forest vegetation types: disturbed lowland forest, disturbed sub-montane forest (hereafter called inhabited lowland forest and cultivated sub-montane forest respectively), montane forest and montane scrub. We also examine the seasonal variation in species richness of the bat community in Mount Cameroon, an area of great importance for global biodiversity conservation.

Material and methods

Study site

Mount Cameroon is a Hawaiian type active volcano that lies on the coast of the Bight of Biafra in the Gulf of Guinea Highlands, Southwest Cameroon (Figure 1). It is the highest mountain in West-central Africa and the fourth most-prominent peak in Africa, rising from sea level to 4,100 m at the summit. It has a surface area of about 1,500 km² (Tchouto 1995), and a long axis 45 km long and 30 km wide running from SW to NE between latitudes 3°57' to 4°27'N and 8°58' to 9°24' E (Suh et al. 2003). According to IUCN 1994, the area is the 10th most conservable places in the world and the most diverse ecosystem in Cameroon. Thomas and Cheek (1992), classified the montane forest of Mount Cameroon into four main vegetation types following an altitudinal zonation:

Figure 1 Map of Cameroon showing location of Mount Cameroon, and sites sampled for bats from November 2016 to July 2018.

Lowland rainforest (0–800 m a.s.l.): it is characterized by trees that vary from 25 m to 35 m in height. The common tree species are Bosquiea angolensis, Kyllinga nemoralis, Alstonia boonei, Microdesmis puberula, Mussaenda elegans, Andropogon tectorum, Justicia tenetla, Fagara macrophylla, and Phymatodes scolopendri, all are often associated with lianas such as Landolphia spp and non-woody climbers such as Melothria spp (Hall 1973). Unfortunately, most of the forest has been cleared and transformed into human settlement, commercial plantations of mainly oil palms (Elaeis guineensis), rubber (Hevea brasiliensis), banana (Musa spp), tea (Camellia sinensis) and cultivated farmland (Payton 1993). The anthropogenic transformation also includes the cultivation of fruit trees such as Carica papaya, Mangifera indica, Psidium guajava, Artocarpus altilis which produce fruits eaten by fruit bats. Also, this area has most of the few water bodies such as springs, streams, rivers, and ponds that provide high quality foraging sites for most bats.

Sub-montane forest (801–1,600 m a.s.l.): is an evergreen covered forest with trees that vary from 20 m to 25 m in height. It is very rich in bryophytes, ferns, and vascular epiphytes and possesses patches of meadows and scrubland dominated by tall Acanthaceae, tall herbs with scattered shrubby trees and tree ferns. This zone is characterized by abundant farmlands, and a part of this forest falls within the zone describe by Hall (1973) as the zone of periodic shifting cultivation. The most common species cultivated are Xanthosoma spp and Dioscorea spp.

Montane forest (1,601–1,800 m a.s.l.): an open forest with fairly discontinuous canopy of medium sized trees up to 20 m in height, dense epiphytes cover and few climbers. The montane forest is drier and receives less rainfall than the forest below 1,600 m altitude, as such it is more susceptible to damage by anthropogenic fires (Forboseh et al. 2011).

Montane scrub (1,801–2,400 m a.s.l.): poorly developed open forest with small sized trees up to 15 m in height, and open layer of small shrubs and herbs. The montane scrub leads to an ecotone that marks the transition between montane forest and sub-alpine grassland beyond.

Aware of threats to the conservation of Mount Cameroon's rich biodiversity, and need to support the development of local communities, the government of Cameroon in December 2009 created the Mount Cameroon National Park (4.055°–4.378°N and 9.031°–9.294°E). The park has a total surface area of 58,178 hectares and encompasses the lowland, sub-montane, montane forests, montane scrub, and montane grassland.

The climate of the Mount Cameroon area is described as humid tropical with, a dry season that extends from November to May and a rainy season that occurs between June and October. According to Proctor et al. (2007), the climate of the area is influenced by closeness to the Atlantic Ocean which causes considerable orographic rainfall. The annual rainfall varies between 2,085 mm and 10,000 mm, and most of the rain occurs between June and September (Cable and Cheek 1998). Rainfall decreases with altitude from 4,000 mm at 1,000 m to about 3,000 mm above 2,000 m, with the summit receiving less than 2,000 mm (Payton 1993). The average annual temperature is about 25 °C, and temperature decreases by 0.6 °C per 100 m rise in altitude (Fraser et al. 1998). Relative humidity is usually high and ranges from 75 to 80% (Watts 1994). The fertile soils of Mount Cameroon are principally of recent origin, mostly on young volcanic rock and have a poor water retaining capacity. This fertile nature of the soil is one of the reasons for the destruction of most of the lowland forest and the establishment of large agro-industrial plantations. The hydrological network consists of a few rivers mostly located at lower elevations. The abundant precipitations and considerable variation of climatic conditions from the bottom to the top of the mountain has an effect on the vegetation type (Payton 1993).

Bat altitudinal sampling

Bats were sampled along an elevational gradient spanning the entire forested cover of Mount Cameroon from sea level to 2,400 m a.s.l. during 20 months of fieldwork from November 2016 to July 2018. Four different elevational ranges that correspond to the natural forest vegetation types on mount Cameroon as described by Thomas and Check (1992) were targeted: inhabited lowland forest (0–800 m a.s.l), cultivated sub-montane forest (801–1,600 m a.s.l.), montane forest (1,601–1,800 m a.s.l.) and montane scrub (1,801–2,400 m a.s.l.). We established 14 different sites within the four forest vegetation types. Sampling sites were selected based on their accessibility and to include different habitat types. Sites one to four, were located in the inhabited lowland forest, sites five to nine in the cultivated sub-montane forest, sites 10 and 11 in the montane forest and sites 12 to 14 in the montane scrub (Table 1). The elevation and coordinates of the different sampling sites were recorded using a hand-held GPS (Garmin eTrex 10). Sampling was carried out for 49 nights comprising 21 nights in the dry season and 28 nights in the rainy season.

Table 1 Sites sampled, characteristics of habitats sampled, number of individuals captured (N) and species richness (n) in the four vegetation types of Mount Cameroon, sampled from November 2016 to July 2018.

Vegetation types	Sites	N	n	Main habitat sampled	Elevation (m a.s.l.)	Longitude North	Latitude East
ILF	1	71	8	Farmland with stream, fruit tree	18	N04°07'47.3”	E009°22'03.6”
	2	14	5	Fruit tree, farmland with stream	21	N04°04'50.4”	E 009°22'03.6”
	3	59	10	Farmland, fruit tree, over a stream	475	N04°11'32.2”	E009°19 03.8”
	4	57	7	Cultivated farmland, fruit tree	791	N04°07'51.5”	E009°14'08.8”
CSMF	5	62	7	Fallow farmland, forest fragments	1,140	N04°08'10.8”	E009°12'34.2”
	6	52	5	Fragmented secondary forest , cultivated farm	1,296	N04°08' 19.0”	E009°11'57.5”
	7	31	3	Cleared farmland	1,311	N04°08' 39.7”	E009°12'17.9”
	8	11	3	Cultivated farmland, forest patches	1,405	N04°08' 50.7”	E009°12'09.5”
	9	13	4	Understory canopy of fruit tree	1,592	N04°09' 00.8”	E009°11'39.9”
MF	10	38	3	Understory canopy	1,645	N04°08' 39.7”	E009°12'09.5”
	11	5	1	Understory closed canopy	1,785	N04°09' 06.1”	E009°11'13.4”
MS	12	7	2	Dense understory canopy, patches of meadows	1,832	N04°08' 41.0”	E009°10'42.3”
	13	6	2	Understory canopy	2,079	N04°08' 50.5”	E009°10'18.3”
	14	140	4	Ecotone forest/grassland	2,242	N04°08' 48.6”	E009°10'10.3”

Abbreviations: Vegetation type=ILF, inhabited lowland forest; CSMF, cultivated sub-montane forest; MF, montane forest; MS, montane scrub.

At each capture site, we placed mist nets in potential areas where bat occurrence was expected, such as around fruit trees, over or near water bodies, cultivated farmland and gaps between tall trees. Mist nets were set just before sunset and monitored every 15 min, till midnight. In order to standardize the sampling effort four mist nets (12 m × 2.5 m; mesh, 40 mm Ecotone-Poland) were deployed at each sampling site per night. After removal from mist nets, bats were identified, measured, weighed, sexed, and photographed with a digital camera (Nikon Coolpix A900) and reproductive condition and age class determined. All bats captured were put in a small bag for fecal collection. Species were identified based on information published in the identification keys of Rosevear (1965), Hayman and Hill (1971), Patterson and Webala (2012), Happold and Happold (2013). Juveniles were distinguished by the incomplete ossification of the epiphyseal plate. Bats were weighed with a portable scale (Ohaus CL501-500 g × 0.1 g), forearm length was measured with dial caliper (Ecotone-Poland 150/0.1 mm). Representative of each species and species that could not be identified in the field were collected as voucher, and preserved in 75% alcohol and deposited at the Biological Laboratory of the University of Maroua. All other individuals were released at the point of capture. Feces of insectivorous bats were preserved in Eppendorf tubes containing 95% alcohol for future laboratory analysis.

Capture, handling and preservation techniques were in accordance with standard guidelines for bats (Kunz and Kurta 1988; Jones et al. 1996; Barlow 1999). The nomenclature was based on Happold and Happold (2013).

Data analyses

In order to test for distributional pattern of bat species on the montane forest of Mount Cameroon, capture sites where spread across four elevational ranges that constitute the different forest vegetation types of Mount Cameroon. The sampling effort was calculated according to Moreno and Halffter (2000) (Table 2), and a Krustal–Wallis test followed by a post-hoc pairwise Mann–Whitney U test was used to assess if species abundance differed among the different forest vegetation types.

Table 2 Individual per species captured at each vegetation type/elevational range of Mount Cameroon, sampled from November 2016 to July 2018.

Species	Inhabited Lowland forest (0-800m a.s.l.)	Cultivated Sub-montane forest (801-1,600m a.s.l.)	Montane forest (1,601-1,800m a.s.l.)	Montane scrub (1,801-2,400m a.s.l)	Total
Pteropodidae
Eidolon helvum	9	2	0	0	11
Epomops buettikoferi	1	0	0	0	1
Epomops franqueti	64	14	0	0	78
Lissonycteris angolensis	6	107	31	26	170
Megaloglossus woermanni	25	30	0	1	56
Micropteropus pusillus	55	2	0	0	57
Myonycteris torquata	3	4	0	0	7
Nanonycteris veldkampii	6	0	0	0	6
Rousettus aegyptiacus	19	2	6	119	146
Rhinolophidae
Rhinolophus landeri	0	1	0	0	1
Hipposideridae
Hipposideros gigas	2	0	0	0	2
Hipposideros ruber	1	2	5	6	14
Molossidae
Chaerephon major	3	0	0	0	3
Mops nanulus	1	0	0	0	1
Mops thersites	1	0	0	0	1
Miniopteridae
Miniopterus fraterculus	0	5*	0	0	5
Vespertilionidae
Glauconycteris egeria	0	0	1	0	1
Hypsugo eisentrauti	0	0	0	1	1
Neoromicia tenuipinnis	4	0	0	0	4
Scotophilus dinganii	1	0	0	0	1
Total	201	169	43	153	566
Number of species	16	10	4	5	20
Capture nights	12	13	8	16	49
Sampling effort (M²h)	3456	3744	2304	4608	14112
Diversity( H')	1.89	1.64	0.84	0.69

*Bats taken from a cave.

The relationship between abundance and species richness was tested using Pearson’s test. Shannon diversity index (H′) and relative abundance was calculated for each vegetation type (Magurran 1988). We classified insectivorous bats into four foraging guilds: aerial insectivores in uncluttered space, aerial or trawling insectivores in background-cluttered space, aerial insectivores in highly cluttered space and gleaning insectivores in highly cluttered space (Schnitzler and Kalko 2001). A Mann–Whitney U test for pair comparisons was performed to test whether there was statistically significant difference between captures in the dry and the rainy season.

Correspondence Analysis (CA) based on presence-absence of species was employed to determine habitat-preferences relationship for most species. Possible association between altitude and bat species richness and abundance were assessed using Pearson’s or Spearman’s correlation depending on data distribution normality. The software packages Microsoft Excel (2013), IBM SPSS statistics for windows, version 20.0 and PAST version 3.0 were used to perform these analyses.

Bat species richness on Mount Cameroon was estimated based on first-order jackknife (Jackknife-1) estimator (Colwell and Coddington 1994), which enabled us to assess the completeness of the surveys by calculating the percentage of estimated species richness that was effectively covered by our sampling. Also, the Mao Tau’s curve of species accumulation with 95% confidence intervals (CI) was plotted for the total number of sampled sites during the study period in order to assess the efficacy and completeness of the study (Figure 2). The software Estimates, version 9.0 (Colwell 2004) was used to perform these analyses.

Figure 2 Species accumulation curve based on the number of bats captured at all 14 sites sampled on Mount Cameroon from November 2016 to July 2018.

Results

In total, the 49 sampling nights yielded 566 individuals belonging to 20 species, 17 genera and six families (Table 2). The number of species, sampling nights and capture effort differed across the different montane forest vegetation types (Table 2), also more captures occurred during the rainy season as compared to the dry season (Table 3). The Shannon diversity index indicated that bat communities were more diverse in habitats under greater anthropogenic influence than in the montane and scrub primary forests (Table 2). The family Pteropodidae was the most abundant, contributing 93.99% of bats fauna of Mount Cameroon. The family Hipposideridae was the second most abundant with 2.83% of the total species.

Table 3 IUCN Red List Status, trophic guild, number of bats captured during the dry and rainy seasons and relative abundance of bats on Mount Cameroon, sampled from November 2016 to July 2018. Abbreviations: IUCN Red list status = LC, least concern; NT, near threatened; DD, data deficient. Trophic guild = F, frugivore; N, nectarivore; HCS, aerial insectivore in highly cluttered space; BCS, aerial insectivore in background cluttered space; US, aerial insectivore in uncluttered space.

Species	IUCN		Captures
	Red List Status	Trophic guild	Dry	Rainy	Total	Relative Abundance (%)
Pteropodidae
Eidolon helvum	NT	F	2	9	11	1.94
Epomops buettikoferi	LC	F	1	0	1	0.17
Epomops franqueti	LC	F	19	59	78	13.78
Lissonycteris angolensis	LC	F	73	97	170	30.03
Megaloglossus woermanni	LC	N	10	46	56	9.89
Micropteropus pusillus	LC	F	10	47	57	10.07
Myonycteris torquata	LC	F	1	6	7	1.23
Nanonycteris veldkampii	LC	N	1	5	6	1.06
Rousettus aegyptiacus	LC	F	112	34	146	25.79
Rhinolophidae
Rhinolophus landeri	LC	HCS	1	0	1	0.17
Hipposideridae
Hipposideros gigas	LC	HCS	0	2	2	0.35
Hipposideros ruber	LC	HCS	6	8	14	2.47
Molossidae
Chaerephon major	LC	US	3	0	3	0.53
Mops nanulus	LC	US	0	1	1	0.17
Mops thersites	LC	US	0	1	1	0.17
Miniopteridae
Miniopterus fraterculus	LC	BCS	5*	0	5	0.88
Vespertilionidae
Glauconycteris egeria	DD	BCS	1	0	1	0.17
Hypsugo eisentrauti	DD	BCS	0	1	1	0.17
Neoromicia tenuipinnis	LC	BCS	0	4	4	0.71
Scotophilus dinganii	LC	BCS	0	1	1	0.17
Total Number of nights			245 21	321 28	566 49	100.00

* Bats taken from a cave

The most abundant species captured on Mount Cameroon was Lissonycteris angolensis (Bocage, 1898) (30.03% of captures), closely followed by Rousettus aegyptiacus (E. Geoffroy, 1810) (25.10%), and Epomops franqueti (Tomes, 1860) (13.78%). Generally, more individuals of each species were captured during the rainy season than dry season except for Rousettus aegyptiacus, with more captured during the dry season (Table 3). A summary of morphometric measurements of all species is presented in Table 4.

Table 4 Ranges and means of morphometric measurements for adult bats (males and females) captured on Mount Cameroon, sampled from November 2016 to July 2018.

Species	N	Sex	FA	TiL	HBL	TaL	TrL	EL	HFL	GLS	WT
Pteropodidae
Eidolon helvum	5	♂♂	103.6-124.1 114.1	42.5-52.2 48.7	147.1-187.5 174.1	6.3-15.6 10.7	- -	26.2-29.2 27.5	28.3-31.4 29.6	48.7-56.0 53.1	140.0-235.0 190.0
Epomops buettikoferi	1	♀	77.2	29.9	97.4	–	–	19.9	18.3	36.1	47.2
Epomops franqueti	24	♂♂	76.9 - 96.0 86.2	29.1-41.2 33.4	102.2-153.2 129.1	- -	- -	17.6 - 26.6 21.9	16.4-21.2 18.8	35.3 - 48.8 43.5	53.0 -142.0 86.1
	42	♀♀	70.9 - 93.0 85.7	28.2-37.2 33.3	104.6-150.4 124.5	- -	- -	19.2 - 25.2 21.5	13.3-24.5 17.9	37.7- 48.8 43.1	35.0-117.0 84.2
Lissonycteris angolensis	18	♂♂	69.0-80.4 76.1	30.0-34.5 32.3	96.3-127.2 133.2	5.5-9.8 7.4	- -	17.1-25.9 19.7	16.2-20.3 18.6	36.4-43.5 40.4	57.4-80.5 66.0
	71	♀♀	71.0-84.7 76.9	28.4-34.6 32.0	100.0-125.2 114.0	3.2-13.2 7.7	- -	14.4-25.1 18.8	13.5-21.5 18.2	35.7-43.7 40.7	53.9-75.8 65.1
Megaloglossus woermanni	22 16	♂♂ ♀♀	39.9-43.4 41.9 38.7-45.5 42.9	17.3-19.4 18.4 16.1-19.8 18.7	61.2-16.1 69.1 64.2-77.2 70.4	- - - -	- - - -	12.5-16.1 14.1 11.9-16.2 14.2	9.2-11.6 10.3 9.1-11.0 10.3	25.4-28.2 27.1 23.2-29.5 27.2	12.5 -15.4 13.5 11.7-15.9 13.5
Micropteropus pusillus	28	♂♂	44.3-54.8 50.1	16.6 -24.8 20.7	63.0 - 88.9 77.2	0-3.8 0.3	- -	10.2-16.5 13.6	10.1-14.9 12.1	23.2-46.4 28.6	15.0-41.0 26.5
	21	♀♀	40.1-55.1 51.6	20.4-24.2 22.5	73.7-89.0 82.9	0-2.3 0.2	- -	11.0-16.3 14.5	11.5-15.2 13.8	24.9-30.4 28.4	20.0-50.0 30.7
Myonycteris torquata	3	♂♂	58.4-63.5 60.5	22.5-24.9 23.5	88.5-97.9 94.1	3.4 -6.9 5.8	- -	16.0-17.0 16.4	13.2-15.6 14.24	31.1-38.8 32.1	30.0-35.0 32.0
	4	♀♀	60.0-62.9 61.2	23.9-25.6 24.8	91.4 - 99.2 94.0	4.5-7.5 6.4	- -	13.5-18.0 17.0	11.7-13.0 13.4	23.5-26.0 31.9	11.0 -20.0 33.5
Nanonycteris veldkampii	3	♂♂	44.8-47.0 46.0	17.1-19.6 18.2	66.2-69.8 68.0	- -	- -	13.9 -16.9 15.4	8.5-12.1 12.5	24.5-28.0 25.0	9.0 -15.0 17.0
	3	♀♀	44.5-49.0 46.4	16.9-20.5 19.4	64.0-73.0 69.2	- -	- -	10.5-15.2 12.8	10.0-11.0 10.5	22.9-26.0 24.4	15.0-23.0 18.8
Rousettus aegyptiacus	105	♂♂	79.2-112.6 91.7	24.8-49.2 41.3	108.0-153.0 129.2	7.9-18.0 12.1	–	15.1-24.8 20.9	17.0-24.8 20.8	37.1-49.2 43.1	73.0-167.0 108.8
	6	♀♀	84.6-101.8 91.1	37.9-44.0 41.2	124.5-140.5 130.4	9.2-13.7 11.9	–	18.3-22.3 20.2	20.7-25.4 22.0	40.3-46.3 43.0	79.0-149.0 105.5
Rhinolophidae
Rhinolophus landeri	1	♂	47.7	19.6	50.1	24.6	11.1	22.6	7.3	20.1	5.9
Hipposideridae
Hipposideros gigas	2	♀♀	107.3-107.9 107.6	43.7-119.3 43.9	110.4-119.3 114.9	34.4-41.8 38.1	- -	26.6-31.2 28.9	13.0-20.1 16.6	35.2-37.9 36.6	97.2-101.5 98.5
Hipposideros. ruber	6	♂♂	49.3-53.6 51.3	19.1-22.2 20.8	49.9-56.3 51.6	26.7-34.5 30.2	- -	12.6-16.9 14.7	5.6-8.4 6.9	14.8-19.8 17.4	5.1-7.4 6.7
	3	♀♀	52.3-54.4 53.0	21.6-22.1 21.9	51.8-55.6 53.4	34.1-34.6 34.4	3.1 -	13.5-15.9 14.9	6.7-7.9 7.2	16.1-18.9 17.9	5.1-9.3 7.0
Molossidae
Chaerephon major	3	♂♂	42.0 - 42.8 42.4	12.1-18.5 14.5	55.0-66.2 63.0	25.0-39.8 32.8	–	9.0-12.2 10.0	5.5-6.7 6.0	16.1-19.5 18.3	14.9-16.7 15.3
Mops nanulus	1	♂	29.9	11.9	54.9	21.7	5.4	14.0	6.5	19.2	8.9
Mops thersites	1	♀	38.5	14.8	57.1	35.6	4.4	16.3	8.8	20.6	7.8
Miniopteridae
Miniopterus fraterculus	1	♂	43.0	17.3	46.8	48.5	4.8	9.5	8.0	15.5	8.7
	4	♀♀	43.1- 44.9 44.0	16.5-17.8 17.4	46.1-50.2 48.2	44.2-47.6 45.9	3.9-4.6 4.2	9.0-12.0 10.3	8.2-8.7 11.3	13.9-16.5 14.8	5.5 -10.2 7.6
Vespertilionidae
Glauconycteris egeria	1	♂	38.5	18.1	43.7	45.1	4.8	11.3	6.3	13.5	4.1
Hypsugo eisentrauti	1	♂	35.4	13.3	44.3	36.5	5.3	11.7	7.2	13.9	3.9
Neoromicia. tenuipinnis	1	♂	28.8	11.1	39.2	27.0	4.7	10.3	4.3	13.2	2.7
	3	♀♀	28.2-30.7 29.3	10.4-11.4 10.9	36.2-40.2 38.9	25.5-28.3 27.2	4.0-5.1 4.5	9.9-12.0 10.8	5.1-5.2 5.2	12.0-13.7 12.9	1.9-2.4 1.8
	4	♀♀	43.1- 44.9 44.0	16.5-17.8 17.4	46.1-50.2 48.2	44.2-47.6 45.9	3.9-4.6 4.2	9.0-12.0 10.3	8.2-8.7 11.3	13.9-16.5 14.8	5.5 -10.2 7.6
Scotophilus dinganii	1	♂	52.8	22.1	70.0	46.8	–	12.2	–	–	27.2

Abbreviations: n = number of bats, FA = forearm length, TiL = Tibia length, HBL = head and body length, TaL = tail length, TrL = tragus length, HFL = hind foot length, GLS = greatest length of skull (all in mm), and WT = weight in grams.

The inhabited lowland forest (0–800 m a.s.l) recorded the highest diversity (16) and individuals captured (n = 201), this was followed by cultivated sub-montane forest (801–1,600 m a.s.l.) with 10 species and 169 individuals. The montane forest (1,601–1,800 m a.s.l.) and montane scrub (1,801–2,400 m a.s.l.) were both low in diversity, four and five species each, and more individuals were capture in the montane scrub as compared to montane forest (43 vs. 153, respectively). The Jackknife-1 species richness estimator suggested a total of 29.29 bats in the area.

The majority of bats captured on Mount Cameroon were frugivorous bats. They were encountered across all vegetation types (Figure 3). Lissonycteris angolensis and Rousettus aegyptiacus were recorded in all vegetation types from sea level to 2,400 m a.s.l and represented 55.83% of captures. Epomops franqueti, Micropteropus pusillus (Peters, 1868), Eidolon helvum (Kerr, 1792), Myonycteris torquata (Dobson, 1878) and Nanonycteris veldkampii (Jentink, 1888) were recorded only at lower altitudes. Most insectivorous bats were recorded at lower elevations except Hipposideros ruber (Noak, 1893), Hypsugo eisentrauti (Hill, 1968), and Glauconycteris egeria Thomas, 1913. Hipposideros ruber was the only insectivorous bat recorded across all vegetation types, it was also the most common insectivorous bat, representing 2.47% of all captures.

A Kruskal–Wallis test revealed that the mean species abundance of bats differed statistically significantly among the four montane vegetation types, (H (3) = 15.12, p < 0.05), and post-hoc Mann–Whitney U test for paired comparisons indicated that species abundance at the inhabited lowland forest differed statistically significantly to that in the montane forest (p = 0.01), and montane scrub (p = 0.03), meanwhile the other vegetation types did not differ significantly from each other (at least p = 0.070). Species richness decreased with altitude but species richness was positively correlated with abundance (r = 0.786, n = 4, p = 0.214). Altitude was negatively correlated with abundance (r_s = –0.800, n = 4, p = 0.200). Number of bats captured did not vary statistically significantly between seasons (Mann–Whitney U test, U = –0.770, p = 0.442).

The graph of dimension 1 × dimension 2 of the correspondence analysis biplot showed a good separation of bat species between the four vegetation types. There was a marked separation between species that characterize zones under anthropogenic influence and those that characterize high altitude undisturbed forest (Figure 4).

Discussion

Studying elevational gradients of montane communities can provide essential information on how communities are restricted by environmental conditions, and enable ecologist forecast the possible effects of climate change on biotic communities (Sundqvist et al. 2013). However, knowledge on montane ecosystems remain limited, particularly information about montane bat communities (Chaverri et al. 2016). The lack of information about montane bat communities can partly be attributed to the challenging nature of the fieldwork, together with the fact that bats are difficult to catch, often resulting in lower sampling efficiency (Alberdi et al. 2013). Patterns of diversity and distribution of montane bats are essential in tackling extinction risk under different climate change scenarios. Indeed, knowledge on altitudinal patterns can improve our capacity to initiate management strategies in time to counter species loss and attenuate impact of climate change on biodiversity, hence are very important in the conservation of biodiversity (Brodie et al. 2012; Stein et al. 2013).

Furthermore, evaluating bat species composition in the forest vegetation types can help elucidate how ongoing anthropogenic alterations can influence local bat assemblages. Indeed, habitat modification by human activities is the principal cause of biodiversity decline worldwide (Ramalho et al. 2014). The rich biodiversity of Mount Cameroon is under threat from anthropogenic activities, mainly poaching, and deforestation caused by agricultural practices such as uphill shifting cultivation, slash-and-burn farming, the extension of agro-industrial plantations (Payton 1993), and the overexploitation of Prunus africana bark that causes tree-die offs (Cunningham et al. 2002). These land-use changes are likely to be detrimental to bats that are forest-dependent, cluttered-space specialist in favor of more open-space generalist bat species (Devictor et al. 2008). As pointed out by Threlfall et al. (2012), bats species that prefer cluttered habitat for foraging are more vulnerable to population decline with increase in deforestation due to a decline in available resources such as suitable roost sites and food. It is possible that ongoing deforestation in the Mount Cameroon area has resulted to a change in the area’s bat assemblage. For example highly cluttered-space insectivorous bats such as nycterids, and forest-dependent frugivorous bats such as Hypsignathus monstrosus H. Allen 1861, Scotonycteris zenkeri Matschie, 1894, and Casinycteris ophiodon (Phole, 1943) were not captured during these recent surveys, despite their presence in the area previously mentioned (Eisentraut 1973; Fedden and Macleod 1986). Thus recent knowledge of local bat fauna is essential, not only for the understanding of regional pattern of bat diversity, but also because bats are useful bioindicators, that show a measurable response to human-induced environmental changes, that can enable ecologist understand the state of health of an entire ecosystem, by comparing species assemblages across different habitats in space and/or time (Medellin et al. 2000; Jones et al. 2009).

In this study, we documented the diversity and distribution of bats along an elevational gradient in the montane forest of Mount Cameroon. The data obtained indicate that the distribution of local bat communities is not homogeneous across the elevational gradient. Bat species richness decreased with elevation and a change in vegetation type (Table 2). Similar studies on tropical elevations outside the African continent followed a similar pattern of increase bat diversity and capture success at lower altitudes (e.g. Patterson et al. 1996; Martins et al. 2015). In Africa, few studies have been conducted on elevational distribution of bats (e.g. Juste and Perez 1995; Curran et al. 2012; Linden et al. 2014; Weier et al. 2016; Reardon and Schoeman 2017). At Bioko Island Equatorial Guinea, Juste and Perez (1995) found that the diversity of fruit bats decreased with increasing elevation. At the southern aspect of the Soutpansberg Mountains of South Africa, Linden et al. (2014) also found a decline in species richness with increasing elevation. Weier et al. (2016) reported a mid-elevation peak in species richness (hump-shaped distribution) on the northern aspect of the Soutpansberg Mountains, whereas species richness declined with elevation on the southern aspect. In contrast, Curran et al. (2012) at Mount Mulanje in Malawi found a low plateau pattern, with high species richness across low elevations, peaking at 1,220 m and thereafter a decline in species richness. More recently, Reardon and Schoeman (2017) studied species richness, functional diversity, and assemblage structure of insectivorous bats along a 1,100 m elevational range in the Mount Nimba area. They concluded that species richness was significantly negatively correlated with elevation and functional diversity declined significantly only at the highest elevation.

McCain (2007b) attributed the high bat species richness at lower altitudes to the increase in the availability of both water and temperature. Indeed, it is well established that calm water bodies with open surfaces provide drinking sites and foraging areas for aquatic emergent insect (Fukui et al. 2006; Korine et al. 2016), resulting in a potential clusters of individuals of different species. Furthermore, temperature influence bat species richness indirectly through availability of food resources: lower temperatures at high altitudes leads to a decline in the availability of insect and plants that produce nectar and fruit (Loiselle and Blake 1991; Soriano 2000). This causes a decline in the number of insectivorous, nectarivorous and frugivorous bats respectively, due to scarcity of food resources. In general, model selection identified vegetation type, which is associated with altitude to be the main factor that influences species richness (Linden et al. 2014).

A species accumulation curve of the number of bat species along the elevational gradient plotted against number of sites did not reach an asymptote (Figure 2), suggesting that not all species were recorded during this study and approximately nine species were unaccounted for. An increase in sampling effort using a variety of capture methods can lead to capture of more species (Colwell et al. 2004). The Jackknife-1 species richness estimation of 29.29 bats on Mount Cameroon, indicates that our surveys recorded approximately 66.9% of the local bat fauna of the area.

According to our results the family Pteropodidae represented the highest number of individuals and their distribution differed according to vegetation type (Table 2, Figure 3). The graph of corresponding analysis biplot displays separation of species among the vegetation types (Figure 4). Dimension 1 separate species that prefer lower altitudes, characterized by greater anthropization on one hand and those that prefer undisturbed primary forest at high altitudes on the other hand. The species that prefer low to mid-altitude elevations below 1,600 m a.s.l., with disturbed lowland and sub montane forest include pteropodids such as Eidolon helvum, Epomops franqueti, Micropteropus pusillus, Myonycteris torquata, and Nanonycteris veldkampii. Epomops franqueti was recorded at elevations between 18 m and 1,405 m and was conspicuously absent at higher altitudes. Eisentraut (1973) and Fedden and Macleod (1986) also recorded Epomops franqueti in disturbed forest below 1,000 m at Mount Cameroon, Mount Kupe and the Rumpi Hills. Eidolon helvum, Micropteropus pusillus, and Myonycteris torquata were characteristic of cultivated and inhabited areas below 1,296 m and were not recorded beyond this elevation (Table 2). This result contrasts with Eisentraut (1973) who documented Micropteropus pusillus at an altitude up to 1,800 m at Mount Manengouba.

Figure 3 Bat abundances in four forest vegetation types of Mount Cameroon, sampled from November 2016 to July 2018. Abbreviations: ILF, inhabited lowland forest; CSMF, cultivated submontane forest; MF, montane forest; MS, montane scrub.

Figure 4 Correspondence analysis scattered biplot based on presence-absence data showing the association between bat species and vegetation types on Mount Cameroon.

Eidolon helvum also preferred lowland disturbed habitats below 1,140 m where it forms large colonies around human habitations. On Mount Mulanje, Malawi, Eidolon helvum was also found in lowland forest up to 1,320 m (Curran et al. 2012), while at Liberian Mount Nimba Eidolon helvum was recorded at 770 m (Verschuren 1976). Our study recorded Nanonycteris veldkampii only at altitudes below 475 m. This result contrast with that obtained at Mount Nimba by Verschuren (1976) which showed that this species had no altitudinal preference and was recorded in both primary and secondary forest. The lone specimen of Epomops buettikoferi (Matschie, 1899) was captured in a cultivated lowland area around a fig tree at an altitude of 18 m. At Guinean Mount Nimba, Verschuren (1976) recorded this species at elevations between 500 and 1,000 m, while Wolton et al. (1982) captured it on the Liberian side of Mount Nimba and found no marked altitude preference. One probable reason why many fruit bats prefer lower altitudes may be because of the presence of garden fruits cultivated by humans. Webala et al. (2014) suggested that Eidolon helvum roost near human habitations because of the proximity of nearby cultivated garden fruits such as mango (Mangifera indica), guava (Psidium guajava), and pawpaw (Carica papaya). Moreover, the availability and abundance of food sources in an ecosystem are known to influence the density of fruit bats (Hodgkison et al. 2004). Also, the cultivated landscapes which characterize lower altitudes are heterogeneous and provide a variety of habitats suitable for a great number of species including bats (Jeanneret et al. 2003).

When compared with previous altitudinal records in West Africa (Verschuren 1976; Wolton et al. 1982; Denys et al. 2013) our data reveal an apparent extension of the altitudinal range of Lissonycteris angolensis and Rousettus aegyptiacus. The abundance of these two species at high elevations may be due to the existence of numerous caves that provide day roost (Kwiecinski and Griffiths 1999). In fact, most bats forage close to their roost in order to maximize net energy gain (Pyke 1984). Megaloglossus woermanni Pagenstecher, 1885, the smallest pteropodid, was recorded in three of the four vegetation types (Table 2), a possible indication of its ecological plasticity. Our results are consistent with the findings of Juste and Perez (1995) and Wolton et al. (1982) who also captured Megaloglossus woermanni in both primary and secondary vegetation. Hipposideros ruber was the only microbat captured in both secondary vegetation at low altitudes and primary vegetation at high altitudes. Only a single individual each of Scotophilus dinganii (A. Smith, 1833), Mops thersites (Thomas, 1903), Mops nanulus J.A. Allen 1917, Glauconycteris egeria, Hypsugo eisentrauti (Hill, 1968), and Rhinolophus landeri Martin, 1838, were captured during our study probably indicating the spatial rarity of these bats in the area. Lissonycteris angolensis was captured in all vegetation types though it showed preference for cultivated sub-montane and montane forest (Table 2, Figure 4). Rousettus aegyptiacus was also captured in all four vegetation types but had a marked preference for the montane scrub (Table 2, Figure 4). A probable reason why these cave dwelling frugivorous bats were encountered at low elevations far from their high elevation day roost sites may be because of altitudinal migration to take advantage of seasonal fluctuation in fruits availability (McGuire and Boyle 2013).

Our study detected molossids only at low altitudes open-space habitats, this can be attributed to the fact that they tend to avoid cluttered environments, where flight can be made difficult due to the abundance of obstacles (Brigham et al. 1997; Grindal and Brigham 1998). Additionally, few molossids were captured during this study, this can be related to the fact that they are high-flying canopy foragers as such are underrepresented in mist nets placed at ground level (Kalko et al. 1996). Some species of vespertilionids were captured at high altitudes while others at low altitudes. Soriano (2000) also found species of vespertilionids that were adapted to cold montane regions in the tropics.

This study has allowed us to collect information on the distribution of montane bat species. However, the use of mist nets only to sample bats is probably bias because not all bats can be captures using mist nets. As pointed out by Kingston et al. (2003) rhinolophids, hipposiderids, and some vespertilionids can easily avoid capture in mist nets because of their versatile flight and efficient echolocation calls. As such, for exhaustive studies different capture methods such as acoustic methods and harp traps are required to supplement mist-netting (Kalko et al. 1996). Indeed, the species accumulation curve of the number of bat species plotted against number of sites did not reach an asymptote suggesting that more species can be recorded if more thorough sampling is carried out (Colwell et al. 2004).

In conclusion, the result of our study provides vital data on the altitudinal ranges of bats in the montane forest of Mount Cameroon. Bat species were not randomly located along the elevational gradient. The pattern of diversity is driven by the availability of environmental resources such as food, water and the proximity of suitable roost sites. Our research shows that most bat species prefer low elevations characterized by disturbed human settlements, secondary forest, and cultivated areas. These species are probably ecological generalist and are more likely to adapt to any environmental changes (Devictor et al. 2008). Species that prefer high-altitude primary forest are probably forest-dependent specialists which are more likely to be threatened by ongoing climate change and habitat degradation including disturbance of caves (Devictor et al. 2008). However, further studies may be required to better assess the impact of forest degradation on long-term population viability of forest-dependent species on Mount Cameroon. In order to protect bats in general and forest-dependent specialists in particular, we recommend the establishment of a buffer zone between areas reserved for agriculture and hunting activities and the boundary of the Mount Cameroon National Park, in order to prevent the encroachment of human activities into protected areas. We also recommend the legal protection of caves that serve as day roost. Given that our study only involves the use of mist-netting technique, we propose that for more exhaustive surveys, complementary techniques such as harp traps and ultrasonic detectors be used, in order to improve the chances of adding more species, most especially aerial insectivorous bats that are canopy foragers and, therefore, are under-sampled in surveys that are based solely on ground-level mist-netting (Kalko et al. 1996).

Acknowledgments

The author thank lecturers of Department of Animal Biology, Faculty of science, University of Maroua for their advice during fieldwork. I am indebted to two anonymous reviewers for providing helpful comments and suggestions, which significantly contributed to improving the quality of the submitted draft. Villagers of Bokwango also assisted during fieldwork by providing field guides, knowledge about the mountain and some logistical support.

Additional information

Funding

We are thankful to the Cameroon Ministry of Scientific Research and Innovation and the Buea Sub divisional office for granting the research permit and authorization to carry out this research (Ref: 0000011/MINRESI/B00/C00/C10/C14). The research was supported by a grant from Rufford Small Grant for nature conservation, UK (Grant Ref: 19621-1).