Biome transitions as centres of diversity: Habitat heterogeneity and diversity patterns of West African bat assemblages across spatial scales

8/8/2019 Biome transitions as centres of diversity: Habitat heterogeneity and diversity patterns of West African bat assembla

1/20

Biome transitions as centres of diversity: habitat heterogeneity anddiversity patterns of West African bat assemblages across spatialscales

Jakob Fahr and Elisabeth K. V. Kalko

J. Fahr ([email protected]), Inst. of Experimental Ecology, Ulm Univ., DE-89069 Ulm, Germany. E. K. V. Kalko, Inst. of ExperimentalEcology, Ulm Univ., DE-89069 Ulm, Germany, and Smithsonian Tropical Research Inst., Apartado Postal 0843-03092, Balboa, Panama.

It is widely accepted that species diversity is contingent upon the spatial scale used to analyze patterns and processes.Recent studies using coarse sampling grains over large extents have contributed much to our understanding of factorsdriving global diversity patterns. This advance is largely unmatched on the level of local to landscape scales despite beingcritical for our understanding of functional relationships across spatial scales. In our study on West African batassemblages we employed a spatially explicit and nested design covering local to regional scales. Specifically, we analyzeddiversity patterns in two contrasting, largely undisturbed landscapes, comprising a rainforest area and a forest-savannamosaic in Ivory Coast, West Africa. We employed additive partitioning, rarefaction, and species richness estimation toshow that bat diversity increased significantly with habitat heterogeneity on the landscape scale through the effects of betadiversity. Within the extent of our study areas, habitat type rather than geographic distance explained assemblagecomposition across spatial scales. Null models showed structure of functional groups to be partly filtered on local scalesthrough the effects of vegetation density while on the landscape scale both assemblages represented random draws fromregional species pools. We present a mixture model that combines the effects of habitat heterogeneity and complexity onspecies richness along a biome transect, predicting a unimodal rather than a monotonic relationship with environmentalvariables related to water. The bat assemblages of our study by far exceed previous figures of species richness in Africa, andrefute the notion of low species richness of Afrotropical bat assemblages, which appears to be based largely on sampling

biases. Biome transitions should receive increased attention in conservation strategies aiming at the maintenance ofecological and evolutionary processes.

Quantifying and explaining the spatial distribution of lifeon Earth is a central focus of contemporary ecologicalresearch. In most taxa, species richness increases from thepoles towards the equator (Hillebrand 2004). Sincestandardized data collection has been rarely achieved overbroad spatial extents, many studies analyzed drivers ofspecies richness using large sampling units such as griddedrange maps or point records generalized to larger areas

(Lyons and Willig 1999, 2002, Ceballos and Ehrlich 2006,Orme et al. 2006, Davies et al. 2007). Accordingly, thesestudies focused on the regional scale as their underlying datado not account for lacunarity or range porosity, that is anincreasing loss of species with increasing spatial resolution(Hurlbert and White 2005, Hurlbert and Jetz 2007). Thecausative mechanisms driving species richness are still hotlydebated, and some of the conflicting results might be

explained by the scale-dependency of species richness(Rahbek 2005).

A major conceptual advancement has been the recogni-tion that local and regional processes act in concert to resultin a community or, more neutrally defined, a point estimateof overlapping regional species distributions (Ricklefs2004). At regional scales, speciation, extinction, andimmigration create, over evolutionary time, regional species

pools. At local scales, habitat selection and species interac-tions as well as stochastic processes may be important. Topredict species richness in relation to environmentalconditions requires an understanding of the relative con-tribution of these processes along a spatially nestedhierarchy (Ricklefs 1987, 2004, Cornell and Lawton1992, Loreau 2000). The landscape scale connects localand regional scales and thus is of immense interest forstudying patterns and causes of species richness (Bohning-Gaese 1997, Whittaker et al. 2001).

Within ecological time, regional diversity sets the limitfor species richness at the local scale. To identify processesthat determine local diversity, we need to ask how beta

Re-use of this article is permitted in accordance with the Term andConditions set out at http://www3.interscience.wiley.com/authorresources/onlineopen.html

Ecography 34: 177195, 2011doi: 10.1111/j.1600-0587.2010.05510.x

# 2011 The Authors. Ecography# 2011 Ecography

Subject Editor: Douglas A. Kelt. Accepted 10 May 2010

177
http://www3.interscience.wiley.com/authorresources/onlineopen.htmlhttp://www3.interscience.wiley.com/authorresources/onlineopen.htmlhttp://www3.interscience.wiley.com/authorresources/onlineopen.htmlhttp://www3.interscience.wiley.com/authorresources/onlineopen.html


2/20

diversity, or species turnover, links regional and local scales.In Whittakers (1960) multiplicative approach, regionaldiversity (in his terms gamma diversity) is the product ofbeta diversity and local (or alpha) diversity. However, thisapproach does not allow direct comparison of the relativecontribution of these factors because regional and localdiversities are measured as the number of species (or relatedunits that incorporate the abundance of species), while betadiversity is dimensionless. Alternatively, diversities can be

partitioned additively where regional diversity0localdiversity'beta diversity (Lande 1996, Loreau 2000, Veechet al. 2002). This additive approach defines beta diversity asspecies turnover and therefore is well suited to analyze therelative contributions of diversity components across spatialscales (Wagner et al. 2000, Crist et al. 2003, Summervilleet al. 2003, Freestone and Inouye 2006, Veech and Crist2007).

Habitat heterogeneity is considered an important me-chanistic factor driving species richness: only few species arefound in all habitats, hence an increase in habitat typesshould lead to more species when sampled across alandscape (Rosenzweig 1995, Kerr et al. 2001, Tews et al.

2004). Several studies assessed this relationship across largespatial extents and employed variables such as altitudinalrange or number of land cover classes per grid cell as proxiesfor habitat heterogeneity (Kerr and Packer 1997, Rahbekand Graves 2001, Van Rensburg et al. 2002, Ruggiero andKitzberger 2004). Since these studies employed relativelycoarse grain, habitat heterogeneity might have been missedas an explanatory variable of species diversity becauseecologically relevant heterogeneity is likely to be perceivedby organisms at finer grain depending on factors such asbody size and dispersal ability.

We assessed diversity and assemblage structure of bats(Chiroptera) in two largely undisturbed areas in Ivory Coast,

West Africa, and asked which factors drive species diversity

from local to regional scales. We employed a spatiallyexplicit and nested design that ranged from local to regionalscales to account for the influence of spatial grain and extent(Wiens 1989, Whittaker et al. 2001, Ricklefs 2004, Rahbek2005). We analyzed constant sample units (plots) that werehierarchically grouped within landscapes, hence keeping thesample grain invariant while changing the sample focus orarea of inference (Scheiner et al. 2000). The extent of thelandscape scale was chosen to match the dispersal abilities ofour study group. As bats show a broad suite of habitat-related adaptations, most notably in their sensory systems(echolocation, passive listening, vision, and smell) andmorphology (wing shape) (Norberg and Rayner 1987,

Neuweiler 1989, Schnitzler and Kalko 2001, Safi andDechmann 2005), we hypothesized that species richness ofbats should be positively related to environmental hetero-geneity as heterogeneous habitats should offer more nichesthan uniform ones. We differentiated between habitatcomplexity and habitat heterogeneity (August 1983), whereheterogeneity is defined as the horizontal variability orpatchiness of a habitat and complexity refers to thedevelopment of vertical strata within a habitat. In ourapproach, heterogeneity of vegetation types is taken as themost relevant habitat parameter for the majority of batspecies as well as the most commonly used variable inprevious studies (Tews et al. 2004).

Our study was conducted in two contrasting landscapesalong the steep climatic gradient of West Africa, which ischaracterized by the staggered arrangement of biomes thatstretch from the rainforest zone in the south throughvarious savanna types up to the Sahara Desert in the north.Variables such as annual precipitation, actual evapotran-spiration, and net primary productivity decrease along thisS-N-gradient while seasonality increases (Tateishi and Ahn1996, Imhoff et al. 2004). If water-related variables best

explain broad-scale patterns of species richness of animals inthe (sub)tropics (Hawkins et al. 2003), species richness ofbats should monotonically increase from deserts to forests.If habitat heterogeneity drives species richness, one wouldexpect a unimodal gradient with a peak at intermediatelatitudes corresponding to the structurally most heteroge-neous region along the biome transition (Guinea Zone)between forests and savannas (Goetze et al. 2006).

We hypothesized first that species diversity increaseswith habitat heterogeneity through the effects of betadiversity. Second, we postulated a positive relation betweenhabitat complexity and species diversity. Third, habitat typerather than geographic distance should explain diversitypatterns across spatial scales. Fourth, we expected that thestructure of functional groups within a habitat type is not arandom draw from the combined landscape assemblage buta selectively filtered subset. Fifth, we hypothesized that thereputed impoverishment of Afrotropical bat assemblages(Findley 1993) is largely due to sampling biases.

Material and methods

Study sites

We assessed bat assemblages in two areas in Ivory Coast, West Africa: Ta National Park (TNP) and Comoe

National Park (CNP), which are ca 500 km apart. TNP(4550 km2) is located in southwestern Ivory Coast andconstitutes the largest protected rainforest in West Africa inconjunction with the adjacent Reserve de faune du N?Zo(790 km2). The study was carried out in the vicinity of theCentre de Recherche en Ecologie station (CRE; 5850?N,7821?W). The rolling landscape (ca 200 m a.s.l.) consists ofa mosaic of drier and wetter parts. The climate issubequatorial seasonal, with an annual precipitation of18139268 mm in the study area (19781982, 19882002;Ta Monkey Project unpubl.) and two dry seasons: a minorone in July August and a major one from December toFebruary. Floristically, TNP belongs to the Guineo-

Congolian regional centre of endemism (White 1983)and the Western Guinean lowland forests ecoregion(Olson et al. 2001). Our study area was composed of amosaic of evergreen forest on the lower slopes with patchesof deciduous trees on hill tops (Van Rompaey 1993). Apartfrom the clearing around the research station, treefall gapsand a few sparsely vegetated inselbergs, the study area iscovered by a closed canopy.

In TNP, we differentiated between two major forest typesaccording to their physical structure: hill forest (foretseche) on slopes and hill tops vs swamp forest (bas-fond)on the bottom of seasonally flooded valleys. Hill forest wascharacterized by high stature of mature trees and a

178


3/20

comparatively open understorey (shrub and herb layer).Swamp forest had a higher density of smaller trees and adenser understorey, mainly composed of Raphiapalms andMarantaceae. We established six plots arranged in threepairs, where one plot of each pair represented hill forest andthe other swamp forest. Distances between plots were 0.22.8 km (median: 2.1 km), with distances between pairedplots of 0.20.3 km, and distances between plot pairs of 1.02.6 km. Despite the short distance between paired plots,

mark-recapture data showed that very few bats crossed fromone plot to its neighbouring pair (19 out of 844 marked bats[2.3%], and 17.6% of all recaptures [108 individuals]),while recapture rate within plots was high (83 recapturedbats [9.8%], and 76.9% of all recaptures), thus justifying totreat each of the paired plots as an independent sample.

CNP (11 493 km2) is located in northeastern IvoryCoast and represents the largest protected area in thesavanna zone of West Africa. The study was conductedaround the former research station of the Univ. of

Wurzburg (Lola Camp: 8845?N, 3849?W, ca 200 m a.s.l.).The landscape generally is flat but occasionally broken byinselbergs or low rocky outcrops. The climate is of humid

Sudanian type, with an annual precipitation of 10039

173mm in the study area (19932002; Univ. of Wurzburgunpubl.), a single dry season from November to March,and a wet season from April to October. Floristically,the southern portion of CNP belongs to the transitionzone between the Sudanian woodland with abundantIsoberlinia and the Mosaic of lowland rainforest andsecondary grassland (White 1983), which is part of theGuinean forest-savanna mosaic ecoregion (Olson et al.2001). The study region is characterized by a matrixof bush-tree savanna with embedded patches of semi-deciduous forest islands ranging in size from 1 ha toseveral km2. Extensive gallery forests with evergreenelements occur along the main rivers Comoe, Iringou,

and Kongo. The wider stretches of gallery forest and somelarger forest islands structurally resemble rainforest andshow floristic affinities to Guineo-Congolian lowlandforests (Hovestadt et al. 1999). The three main habitattypes (savannas: covering 84.2% of the area; forest islands:8.4%; gallery forests: 2.3%) result in an overall mosaic-likelandscape structure with clearly defined edges betweenvegetation types (Hovestadt et al. 1999, Hennenberget al. 2005, Goetze et al. 2006).

In CNP, we sampled bat assemblages in three vegetationformations: open bush-tree savanna (savanes boisee andsavanes arbustive), forest islands, and gallery forest.Initially, we established two plots in each of these three

habitat types. A third savanna plot was added duringthe second half of the study period, resulting in a total ofseven plots. Distances between plots ranged between1.4 and 13.9 km (median: 5.4 km). The large distancescompared to TNP result from one plot in gallery forest thatwas situated rather far from the other plots.

Sampling design

Each plot comprised 12 mist nets arranged in a standardizedconfiguration along a 200 )100 m-rectangle (2 ha), withequidistant (50 m) centres of the nets. The nets were

oriented in an alternating fashion perpendicular to oneanother. These understorey nets (UN) were set on polesnear ground level or slightly elevated, with the lower netedge level with the surrounding soil or herb layer. Inaddition, we set up one elevated net system in each plot,which consisted of a pulley and rope structure to hoist fourstacked nets usually reaching a height of ca 25 m. Thesecanopy nets (CN) were established within, or close to, therectangle formed by the understorey nets, in TNP within

natural treefall gaps, in CNP either within gaps (forest plots)or between emergent trees (savanna plots). All mist netsmeasured 12)2.8 m (16 mm mesh; 70 denier/2-plynetting) with four or five shelves. Furthermore, one two-bank harp trap (4.2 m2 capture area; Faunatech) was set ineach plot.

Each plot was typically sampled for two consecutivenights per field season. Capturing lasted from dusk (ca18:30) until dawn (ca 06:30). Mist nets and the harp trapwere checked every 3060 min throughout the entire night.

We did not capture during nights around full moon phases,and in rare cases interrupted sampling because of heavy rain.

Bats were measured (forearm, body mass) and their sex,

age, and reproductive status assessed. Most individuals wereidentified to species in the field and subsequently released. A few bats were sacrificed to check identifications. Thesesynoptic collections are deposited in the ForschungsinstitutSenckenberg, Frankfurt/M., and in the research collectionof JF. All adult bats except for insectivores with B10 gbody mass and males of Epomophorus gambianus, Epomopsbuettikoferi, and Hypsignathus monstrosus were individuallymarked with a stainless steel ball-chain necklace and aserially numbered aluminium band. Males of the threespecies of fruit bats were not marked as they inflate theirthroats during courtship calls.

The study comprised eight field seasons in TNP betweenMarch 1999 and February 2004 (first part of the study

[19992000]: J. Fahr; second part [20012004]: StefanPettersson, Goteborg Univ.). In CNP, we sampled batsduring seven field seasons between April 1999 and June2002 (first part of the study [19992000]: J. Fahr; secondpart [20012002]: Njikoha Ebigbo, Ulm Univ.). Captureseasons were selected to match similar conditions inphenology and climate, i.e. at the end of the dry season/start of wet season (TNP: Feb/Mar; CNP: Apr/May) and atthe end of the wet season/start of dry season (TNP: Aug/Sep; CNP: Oct/Nov).

We also captured bats outside plots in an opportunisticfashion with mist nets set in understorey and canopy as wellas with harp traps. Such opportunistic sampling (OS)

targeted particular habitat types and situations (e.g. smallcreeks, clearings, and rocky outcrops), which were deemedto yield species that might have been missed in thestandardized plots. Additionally, we included data frompreliminary surveys in CNP during 1993 and 1995. Totalcapture effort for CNP and TNP combined was 1765.0mist net nights and 102.6 harp trap nights (Table 1).

Data analysis

Recaptures of marked bats were excluded from analysis ifthey were caught during the same sampling period in the

179


4/20

same plot. Estimated species richness (Sest) was calculatedwith the programs EstimateS 7.5 (Colwell 2005). Wefollowed Brose and Martinez (2004) for the choice of theleast biased and most precise estimator to extrapolateestimated species richness, Sest. In a first step, we calculatedSest of a given sample with a suite of non-parametric andparametric estimators of species richness (Abundance-basedCoverage Estimator [ACE], Incidence-based CoverageEstimator [ICE], First-order Jackknife [Jack1], Second-order Jackknife [Jack2], and Michaelis-Menten [M-M]).

We then calculated the range of sample coverage (observedspecies richness [Sobs]/estimated species richness [Sest]) andits mean. Variation in estimated sample coverage was

generally rather low (8

29%). In a second step, we chosethe estimator recommended by Brose and Martinez (2004)as the final estimate of species richness for a given sample.

Interpolated species accumulation curves (sample-basedrarefaction) of plot data were calculated with the MaoTau-function in EstimateS (Colwell et al. 2004, Colwell2005). The graphs were rescaled by individuals, resulting inindividual-based rarefaction curves sensu Gotelli andColwell (2001). Rescaling by individuals allows directcomparison of species richness as opposed to rarefactioncurves scaled by samples, which represent species density(Gotelli and Colwell 2001). We used 95% confidenceintervals to test for significant differences in species richness.

We followed Hill (1973) and Jost (2006) in the use of

effective number of species when reporting diversitymeasures other than species richness. In short, the effectivenumber of species equals species richness if all species of asample have the same frequency and decreases withdeclining evenness of a sample. Shannon diversity, whichis equivalent to Hills (1973) N1 diversity index, wascalculated as eH, with

H 0(XS

i01

pi lnpi

and Simpson diversity was calculated as 1/D, with

D 0XS

i01

p2i

where pi0the proportion of individuals in the ith species.We also employed the nonparametric estimator of Shannonentropy implemented in Spade 3.1 (Chao and Shen 2006),which accounts for unseen species in a sample, thusresulting in Shannon diversities (eH [est]) that are unbiasedby sample size. We stress that frequency data derived fromcaptures represent relative abundances of individuals, whichin turn are affected by sampling bias of capture techniques.Since our sampling protocol was standardized, comparisons

within our study system are valid since data are affected bythe same bias.

Species richness was used to assess the total number ofspecies in a sample (diversity of order 0 sensu Jost 2006),Shannon diversity was employed as a diversity measure thatweighs species directly proportional to their frequencies(diversity of order 1), and Simpson diversity was used as acomplement to focus on the most frequent species in asample (diversity of order 2). Evenness was calculated asE0eH/S, where S0number of species in a sample (Buzasand Hayek 1996). Evenness equals 1 if all species in asample have the same frequency and decreases as samplesare increasingly dominated by a few species, hence reducing

the effective number of species (e

H

). We calculatedobserved evenness from observed Shannon diversity andSobs as well as estimated evenness (E) derived from eH

andSest. For the latter, estimators were chosen based on samplecoverage, see above.

Spatial variation in beta diversityTo analyse whether variation in assemblage compositionamong sites within a region (variation in beta diversity,Tuomisto and Ruokolainen 2006) is explained by geogra-phical location (spatial autocorrelation), we ran indepen-dent Mantel tests for each study region. Plot data (relativespecies abundances per plot) were transformed as dissim-

ilarity matrices based on the quantitative Srensen (Bray-Curtis) index. Geographical locations of plots had beenmeasured with a hand-held GPS (Garmin GPS II plus) andthe Euclidean distances between the midpoints of each plotwere calculated with ArcView 3.2a. We compared bothdistance matrices of each study region, using PC-Ord 5(McCune and Mefford 2006) to run 10 000 Monte Carlorandomizations.

Additive partitioningThe program PARTITION (Veech and Crist 2009) wasemployed to assess additive partitioning of species richnessfor each study region (CNP, TNP). We tested the null

hypothesis that the observed components of diversity atincreasingly higher levels (a1, b1, b2, . . ., bi) could have beenobtained by the random placement of individuals amongsamples at all hierarchical levels (Crist et al. 2003, Crist andVeech 2006). For this approach, the observed numbers ofindividuals of each species are randomly placed amongsamples at the lowest hierarchical level, and these samplesare then grouped into progressively larger samples at eachhigher level. Under this null model, each species is notconstrained to a particular sample (in our case representinga specific habitat type) but reshuffled among samples,thereby effectively removing the influence of species-specificassociations with a particular habitat type (see also Veech

Table 1. Capture effort expressed as mist net nights (UN: understorey nets, CN: canopy nets) and harp trap nights (HT: harp traps). 1 net night:one 12 m-net opened for 12 h, 1 trap night: one trap set for 12 h. Opportunistic sampling in CNP includes UN-data from 1993 and 1995.

Standardized plots Opportunistic sampling

UN CN HT UN CN HT

CNP 538.7 155.7 43.1 91.8 4.9 5.0TNP 512.8 353.4 43.6 63.9 44.0 11.0

180


5/20

and Crist 2007). The program PARTITION placesindividuals randomly in samples while preserving theoriginal species-abundance and sample-size distribution.For randomizations (10000), we arranged our samples(plots) in replicates corresponding to their spatial location inthe landscape, thus matching a nested design. For TNP, thisresulted in three replicates each containing two samplesrepresenting distinct habitats (hill forest, swamp forest). ForCNP, this resulted in two replicates each containing three

samples representing distinct habitat types (forest islands,gallery forests, savannas). Samples were weighted by therelative number of individuals in each sample, i.e. eachsample received a weight equal to the individuals in thissample divided by total number of individuals.

Our analytical design spanned three hierarchical levelsfrom the local to the landscape scale (Fig. 1). Additivepartitioning allows the expression of the proportionalcontributions of diversity at each level in this hierarchy.Since diversities are calculated as an average of the samplesat a given level regardless of how they are nested within thenext higher level, this approach is robust to unbalancedsampling designs (Summerville et al. 2003). Although the

number of structural vegetation types, and therefore oursampling design, differed between CNP and TNP, therelative contributions of diversity components across spatialscales can be compared between both study regions.

Functional group compositionBats were classified into five broad functional groupsfollowing Schnitzler and Kalko (2001) based on diet (frugi-and nectarivorous [F] vs animalivorous [A]), foraging mode(gleaning [g] vs aerial [a]) and habitat (degree of structuralclutter: narrow space [NS; foraging within dense vegeta-tion], edge and gap [EG; foraging close to, but not withindense vegetation], open space [OS; foraging distant from

vegetation]) (Appendix).We assessed whether habitat type structured functional

group composition of assemblages by testing whether theobserved composition in a given habitat type conformed toa random sub-sample of each study area (TNP, CNP) or ifthe proportional composition of functional groups shows

habitat-specific patterns. We employed the programResampling Stats (Resampling Stats 2006) to create10 000 random assemblages for each habitat type. Thesewere constrained by drawing the observed number ofspecies without replacement from the species pool of eachstudy region (TNP: 40 species, CNP: 57 species). Statisticalsignificance was calculated as the proportion of null valuesgreater than (or less than) the observed values. Thisproportion is a p-value that indicates the probability ofobtaining a value as great as (or as small as) the observedvalue by chance.

Results

Landscape diversity and sample coverage

We captured a total of 75 species, 22 of which were sharedbetween the two study areas (Appendix). We recorded 40species in TNP, with 32 species caught in plots (P) andeight species captured opportunistically (OS). The total forCNP was 57 species, with 51 species recorded in plots and

six species that were found off-plot.Standardized plot data revealed significantly higher

species richness for CNP than for TNP when rarefied tothe assemblage with the lower number of individuals; thus,at 1569 individuals, TNP had 39 species (upper 95% CI:42 species) whereas CNP had 50 species (lower 95% CI: 45species; Fig. 2). The completeness of sampling wassimilar for both areas as indicated by high sample coverage(Sobs/Sest CNP: 80.3%, TNP: 78.5%) despite the veryunequal number of individuals captured in each study area(Table 2). Including data from opportunistic sampling,estimated sample coverage increased to 88.191.5% forCNP and 87.795.5% for TNP, respectively. At this level

of sample coverage, the magnitude of the higher speciesrichness of CNP compared to TNP was much morepronounced (TNP: 39 species, upper 95% CI: 42 species;CNP: 55 species, lower 95% CI: 51 species).

mean 1+

mean 1

mean 2

mean 2 +

=

gamma

=

A-C B-C

habitat A habitat B

habitat C

A-B

A-C B-C

habitat A habitat B

habitat C

A-B

2 (A+B+C) 2 (A+B+C)2

(landscape)

Replicate 1 Replicate 2

Figure 1. Schematic representation of the hierarchical levelsstudied in CNP. For TNP, the arrangement differed in that therewere two habitat types (rather than three) and three replicates(rather than two). The right-hand circles illustrate how each lowerlevel adds to the next hierarchical level (g0a1'b1'b2).

Number of individuals

0 500 1000 1500 2000 2500

Numberofspecies

0

10

20

30

40

50

60

70

CNP Jack1 1 SD

TNP Jack1 1 SD

CNP Sobs

TNP Sobs

Sobs 95% CI

Figure 2. Estimated number of species (extrapolated speciesaccumulation: mean Jackknife 191 SD) for CNP (blackdiamonds) and TNP (grey circles), and sample-based rarefactioncurves (interpolated species accumulation: Sobs) with 95% con-fidence intervals rescaled by individuals.

181


6/20

Table 2. Species richness of CNP and TNP broken down to approach (P: plots, OS: opportunistic sampling, pooled: P'OS) and method (UN: understorey net, UN'CN'HT). Species numbers in brackets refer to those captured with a single method. Sobs: observed species richness, Sest: estimated species richness, e

H: Shannon diversity, Eobs: observed evenness, 1/D: Simpson diversity. Sest a: Michaelis-Menten, b: ICE, c: Jackknife 1, d: Jackknife 2. In CNP, one additional speday roost, another additional species was found in its day roost and recorded by its echolocation calls (Rhinolophus landeri); in TNP, one additional species (Mystation building.

Comoe NP Ta

Combined UNpooled

CNpooled

HTpooled

Combined

pooled P OS pooled P OS

Pteropodidae 9 9 9 9 9 1 8 8 7

Emballonuridae 1 1 1 1 (1) 1 1 Nycteridae 5 5 2 5 (4) 1 5 4 3Rhinolophidae 1 1 1 1 1 1 1 1 1Hipposideridae 6 5 5 6 (2) 1 3 6 6 4Vespertilionidae 20 17 13 13 (5) 12 (4) 5 (3) 11 10 6 Molossidae 13 13 4 2 13 (11) 7 2 7

Sobs 55 51 35 36 (11) 38 (16) 10 (3) 39 32 28 Sest 60

a62b 63c 45c 47c 13c 41a44b 41c

eH 12.1 12.2 9.8 11.0 6.0 10.3 8.6 eH 12.3 12.4 10.0 11.4 6.7 10.5 8.8 Eobs (e

H/Sobs) 0.22 0.24 0.27 0.29 0.60 0.26 0.27 1/D 6.5 6.4 6.1 6.1 4.9 5.5 5.0 Individuals 2443 1945 498 1472 904 67 1569 1307 262

182


7/20

Although species richness (Sobs) was much greater usingunderstory and canopy nets than with harp traps, eachmethod (UN, CN, HT) yielded species not documentedwith other methods (Table 2). In CNP, observed Shannondiversity (eH) was highest for the combined samplingmethods (both P and P'OS), decreased slightly for mistnet samples (both understorey and canopy level), and waslowest for harp trap samples. In TNP, Shannon diversitywas highest for mist net samples at understorey level,

intermediate for harp trap samples and combined methods(P and P'OS), and lowest for canopy samples. For mostsamples in both study areas, estimated Shannon diversity(eH ) was only marginally higher compared to observedShannon diversity, showing that the influence of undetectedspecies on Shannon diversity was negligible due to highsample coverage. Only in TNP, understorey net samples(12.4 vs 13.8) and harp trap samples (9.7 vs 11.4) hadnoticeably higher estimated than observed values.

Observed evenness generally was very similar amongmethods and between samples except for harp trap samples(both CNP and TNP) and understorey net samples inTNP, which had a much higher evenness than the

remaining samples (Table 2). Finally, Simpson diversity(1/D), which emphasizes the most frequent species, washighly similar among CNP samples except for HT, whichhad a lower diversity. Simpson diversity in TNP was highfor UN- and HT-samples, intermediate for combinedmethods (P and P'OS), and low for CN-samples.

The estimated total species richness (Sest) of CNP whenbased on a single method (UN, CN, HT) was lower thanthe observed number of species (Sobs) and much lower thanthe estimate derived from all methods combined. In TNP,samples from understorey and canopy nets estimated aspecies total close to the observed as well as estimatedrichness of all methods combined while harp trap samplespredicted roughly two-thirds of the total species richness.

Spatial variation in beta diversity and additivepartitioning of species richness

Variation in community composition among plots wasuncorrelated with geographic distance in CNP (Mantel test:r 00.372, p00.2146) in contrast to TNP (Mantel test:r 00.520, p 00.0067). However, when substituting theaverage community dissimilarity value of TNP-plots for thethree plot pairs (P1P2, P3P4, P5P6) in close proximityto each other, geographic distance no longer had asignificant effect on community composition (Mantel test:

r0

0.183, p0

0.2171). We obtained similar results (notshown) for all tests when applying Jaccard or relativeSrensen distance indices.

At the local scale (plots: a1), additive partitioningrevealed significantly lower species richness both in TNPand CNP compared to the individual-based randomiza-tions, that is, a lower proportion of species richness wasfound on this level than expected from a random distribu-tion of species between plots in both study regions (Table 3,Fig. 3). CNP had significantly higher beta diversity on bothlevels (b1: among plots representing different habitat types,b2: between replicates) compared to the null model ofrandom placement of species. Both beta components of T

able3.AdditivepartitioningofspeciesrichnessforstandardizedplotsinCNPa

ndTNP.Observedvaluescomparedtome

an(min-max)valuesofindividual-basedra

ndomizations(10000)andthe

proportion(p)ofrandomizedvalueswithadiversityestimategreaterthantheobserved.Arrangementofsamplesconformtothe

spatiallynesteddesign(CNP:threeplotseachrepresentingdifferenthabitat

typesonthefirstlevel,whichareaggregatedintothesecondlevel[tworeplicates];TNP:twoplotseachrepresentingdifferenthabitattypesonthefirstlevel,whicharea

ggregatedintothesecondlevel

[threereplicates]).NotethatdiversitypartitionsadduptototalSobs

inplots(a1'

b1'b2;CNP:51,TNP:32).

ComoeNP

TaNP

Observed

diversity

%oftotal

diversity

Expected

diversity

p

Observed

diversity

%oftotal

diversity

Expected

diversity

p

Withinplots(a1)

25.6

50.3%

33.1(30.435.5)

0.9999

18.9

59.0%

20.4

(18.521.8)

0.9996

Amongplots(b1)

15.6

30.5%

11.7(8.514.8)

B0.0001

4.7

14.6%

3.9

(1.95.8)

0.0715

Betweenreplicates(b2)

9.8

19.2%

6.3(3.89.4)

0.0001

8.5

26.4%

7.7

(5.910.1)

0.0866

183


8/20

species richness in TNP did not differ from the null model.In TNP, 41% of total species richness was apportioned onthe beta levels while in CNP the combined contribution ofbeta diversity to total species richness was 50%. The studyareas also differed in the proportional partitioning of betadiversity on the two hierarchical levels of our analysis: inCNP, a larger fraction of observed species richness wasfound on the first level (b1), corresponding to high species

turnover between plots representing different habitat types,whereas in TNP a larger fraction of species richness wasfound on the second level (b2), representing a higher speciesturnover between replicates (paired plots).

Local (alpha) diversity

In CNP, estimated species richness was highest in galleryforests and one savanna plot (CP5), while estimatedShannon diversity was highest in gallery forests, mediumin forest islands, and lowest in savanna plots (Table 4). Inpair-wise comparisons (95% CI), Shannon diversity ingallery forest plots (CP3, CP4) was significantly higher thanin forest island and savanna plots except for one forestisland plot (CP2), which did not differ significantly fromCP3. Estimated evenness was very similar among plotsexcept for CP5, which had a much lower evenness. Despitethe high species richness in savanna plot CP5 (both Sobs andSest), Shannon and Simpson diversities were very low as aresult of the disproportionately high dominance of twofruit bat species (Micropteropus pusillus and Nanonycterisveldkampii), which is also reflected in the much lowerevenness of this plot. In TNP, estimated species richnesswas very similar among plots except for one located inswamp forest (TP5), which was predicted to have muchhigher species richness. Estimated Shannon diversity and

evenness was variable and not consistently related to habitattype (in pair-wise comparisons [95% CI], only TP4 wassignificantly more diverse than TP2). Plot TP5 wasdistinguished by the exceptional dominance of one fruitbat species (Eidolon helvum), which might have beenattracted to food resources nearby. The same plot alsohad a very high number of singletons (7), which resulted ina high Sest.

Overall, species richness of plots was weakly andpositively correlated with Shannon diversity (Fig. 4; linearregression for observed and estimated values, respectively:eH02.230'0.295 Sobs, R

200.207, p 00.137; eH0

2.503'0.230 Sest, R2

00.293, p 00.069). Species richnessand evenness were independent from each other (linearregression for observed and estimated values, respectively:E00.4880.004 Sobs, R

200.026, p 00.617; E00.412

0.003 Sest, R2

00.075, p 00.390). Both observed andestimated mean species richness of CNP-plots was signifi-cantly higher than TNP (t-test, Sobs: mean(CNP) 024.8,mean(TNP) 018.5, t03.463, 10 DF, p 00.006; Sest:

mean(CNP) 035.4, mean(TNP)024.7, t 03.245, 10 DF,p 00.009).There were neither significant differences between TNP

and CNP with respect to Shannon (both observed andestimated) and Simpson diversities nor with regard toevenness of plots. However, potentially significant differ-ences on the level of Shannon and Simpson diversities mighthave gone undetected as a result of low sample size (n 06 forTNP and CNP, respectively). On the level of observedspecies richness, plots in CNP harboured a significantlylower proportion of total species richness compared to TNP(t-test, % of Sobs: mean(CNP)048.7%, mean(TNP)057.8%, t 0 (2.286, 10 DF, p 00.045). However, therewas no significant difference between CNP and TNP when

based on estimated species richness (t-test, % of Sest:mean(CNP)056.2%, mean(TNP) 060.3%, t 0 (0.642,10 DF, p 00.535).

In CNP, forest islands had significantly lower speciesrichness than gallery forests and did not differ fromsavannas when Sobs was rarefied to the smallest sample(GF, Table 5). Although the corresponding values of Sobsdiffered widely between gallery forests and savannas, thesignificance level was just missed due to wide confidenceintervals for both habitats. Estimated species richness (Sest)of gallery forest approached the total estimated for CNPand was much higher than in savannas, which in turn hadhigher Sest than forest islands. Shannon diversity (e

H) was

significantly higher in gallery forests compared to bothforest islands and savannas while no significant differencewas found for Simpson diversity (1/D; note very wideconfidence intervals in the latter). Observed evennessdecreased from gallery forests to forest islands and savannas,whereas estimated evenness was similar among foresthabitats but lower in savannas. Higher evenness in foresthabitat, particularly in gallery forest, was caused by muchlower capture frequencies of the two dominant fruit bats(Micropteropus pusillus, Nanonycteris veldkampii).

TNP-plots grouped by habitat type did not differ in anyof the diversity measures (Sobs when rarefied to the smaller

Observed Expected Observed Expected

Speciesrichness(%o

ftotal)

0%

20%

40%

60%

80%

100%2

1

1

2

1

1

0%

20%

40%

60%

80%

100%

Figure 3. Additive partitioning of diversity components (a1, b1, b2) for CNP (left) and TNP (right) as the percentage of total speciesrichness. Expected values are derived from the means of 10 000 individual-based randomizations (Table 3).

184


9/20

sample[HF],Shannon,Simpson).However,hillforestshad

lower

species

richness

(Sest)

b

uthigher

observed

and

estimatedevennessthanswampforests.

Structure

offunctionalgroups

in

relation

to

habitat

type

In

CNP,thecomposition

off

unctionalgroupsshowed

severalsignificantdeparturesfromthenullmodelofa

random

draw

ofspeciesintoloc

alassemblages(Table6).

Thebatassemblageofforestislandswascharacterizedbya

significantlyhigherrichnessof

frugi-and

nectarivorous

gleaningnarrow

spacebats(FgN

S)andbyasignificantly

lowerrichnessofanimalivorousaerialopen

spacebats

(AaOS).Theassemblageofgall

eryforestsconstitutedan

almostperfectrandom

draw

fro

mthespeciespool.The

savannaassemblageshowedtheg

reatestdeparturefrom

the

nullmodel:richnessofanimaliv

orousaerialnarrow

space

bats(AaNS)wassignificantlylower,buthigherforAaOS

bats;FgNS

batstended

to

be

more

species-rich

than

expected

(p00.0517)

while

t

hefunctionalgroup

of

animalivorousgleaningnarrow

spacebats(AgNS)tended

tobeimpoverished(p00.0854).

In

TNP,the

AaOS

group

exhibited

lowerspecies

richnessbothinhilland

swam

pforestwhiletheFgNS

Table 4. Species diversities of plots in CNP and TNP. Indiv.: individuals, Sobs: observed species richness,% ofSobs (total): percent of total Sobs (CNP: 51, TNPobserved evenness (eHobs/Sobs), 1/D: Simpson diversity, Sest: estimated species richness, Est.: estimator used for Sest (Jack1, Jack2, Jack3),% ofSest (total): percenestimated Shannon diversity, eH [est] SE: standard error of estimated Shannon diversity, Eest: estimated evenness (e

Hest/Sest).

Plot Habitat Observed Es

Indiv. Sobs % of Sobs(total)

eH Eobs 1/D Sest Est. % of Sest(total)

CNP P1 forest island 301 23 45.1% 9.1 0.40 6.3 32.7 Jack2 51.9%CNP P2 forest island 245 23 45.1% 9.9 0.43 6.4 29.0 Jack1 46.0%CNP P3 gallery forest 179 25 49.0% 13.1 0.52 9.6 39.0 Jack3 61.9%CNP P4 gallery forest 209 29 56.9% 14.3 0.49 10.0 44.2 Jack2 70.1%CNP P5 savanna 477 30 58.8% 6.2 0.21 3.3 39.6 Jack1 62.9%CNP P6 savanna 185 19 37.3% 8.0 0.42 5.8 28.2 Jack2 44.7%TNP P1 swamp forest 159 18 56.3% 6.1 0.34 3.0 21.3 Jack1 52.0%TNP P4 swamp forest 189 20 62.5% 9.6 0.48 6.3 24.3 Jack1 59.2%TNP P5 swamp forest 326 21 65.6% 5.3 0.25 3.1 34.4 Jack2 83.9%TNP P2 hill forest 282 18 56.3% 4.8 0.26 2.5 22.4 Jack1 54.6%TNP P3 hill forest 105 16 50.0% 8.4 0.53 5.9 23.8 Jack2 57.9%TNP P6 hill forest 246 18 56.3% 8.8 0.49 6.0 22.3 Jack1 54.4%

Speciesric

hness[S]

16

18

20

22

24

26

28

30

Shannon diversity [eH]

2

Speciesric

hness[S]

20

25

30

35

40

45

2 4 6 8 10

12

14

16

18

CP1

CP2

CP

3

CP4

CP

5

CP

6

TP1

TP4

TP

5

TP2

TP

3

TP

6y=2.5

03+

0.2

30x

R2

=0.2

93

Figure4.LinearregressionbetweenspeciesrichnessandShannon

diversityofplots.Above:observed;below:estimated.(TNP-plots:

TPx,CNP-plots:CPx,Table4).

185


10/20

group had higher species richness than expected by the nullmodel in both habitat types. Most other functional groupsshowed proportional sampling close to that expected fromthe species pool although AaNS bats tended to be morespecies-rich in both habitats compared to the null model,albeit not significantly so (p 00.0612 and 0.0827 for SFand HF, respectively).

At the landscape level, the functional group composi-tion of both assemblages (TNP, CNP) constituted

random draws from the regional species pool recordedfor Ivory Coast (data not shown). We obtained the sameresult when comparing the observed composition of eachstudy area (TNP: 40 species, CNP: 57 species) with arandom draw from the combined species total of TNPand CNP (75 species) or with a random draw from theentire species pool of Ivory Coast (87 species; Fahrunpubl.).

Discussion

We discuss our results in a spatially hierarchical order. First,

we assess patterns of local diversity in relation to habitattype. Second, we evaluate how composition of functionalgroups is a function of differential recruitment from thespecies pools as mediated by habitat type. Third, we showhow patterns on the local level scale up to the landscapelevel through species turnover (beta diversity), which islinked to habitat heterogeneity, and introduce a conceptualmodel that aims to dissect the contribution of habitatcomplexity and heterogeneity along biome gradients.Finally, we discuss our findings in the light of previousstudies on tropical bat assemblages.

Relating local (alpha) diversity to habitat type

We used the contrasting landscape configuration andhabitat types of CNP and TNP to assess the influence ofhabitat complexity on local diversity and found that habitattype had a pronounced influence on local diversity in CNPbut not in TNP. With the exception of one savanna plot(CP5), estimated species richness decreased from galleryforests through forest islands to savannas. Shannon diversityrevealed a similar pattern except for one pairwise compar-ison between forest island and gallery forest (CP2-CP3).These results suggest that local bat diversity is positivelyrelated to complexity along the vertical axis. It remainsunclear, however, why gallery forests supported higher local

diversity than forest islands in CNP. Possibly, the linearstructure and narrow width of gallery forests might havefostered higher permeability for savanna species along theedge.

At the level of species richness, mean single-plot diversitywas significantly higher in CNP than in TNP (i.e. acrosshabitat types). Considering that plots size and samplingwere standardized, local diversity in CNP was possiblyincreased through a spillover of species between differenthabitat types. This could have been partly caused by oursampling design, which deliberately included edge habitat,and may constitute a mass effect (sensu Shmida and Wilson1985) through species occasionally extending their coreTa

ble5.Batdiversityinrelationtohab

itattype(CNPGF:galleryforest,FI:forestisland,SA:savanna;TNPSF:swampforest,HF:hillforest).CI:95%confidenceinterval;Sest

a:Jackknife2,b:ICE,

c:Michaelis-Menten,d:Jackknife1;ra

refied:comparisonrestrictedtothesample

withthelowestnumberofindividuals.

Habitat

Indiv.

Sobs

CI(Sobs)

Sest

eH

eH[est]

CI(eH

)

1/D

CI(1/D)

Eobs

Eest

ComoeNP

GF[P3-4]

388

39

31.846.2

62.1

a

18.0

19.7

15.4

25.2

13.1

9.720.2

0.46

0.32

FI[P1-2]

546

28

26.529.5

30.5

b35.6

c

10.2

10.6

8.0

13.9

6.6

4.611.5

0.36

0.300.35

FIrarefied

390

26.3

24.628.1

SA[P5-6]

662

34

29.438.6

43.6

d

7.5

7.9

5.1

12.5

4.1

2.69.8

0.22

0.18

SArarefied

388

28.1

23.832.4

TaNP

SF[P1,4,5]

674

28

20.135.9

36.5

d

8.1

8.4

6.1

11.6

4.7

3.29.1

0.29

0.23

SFrarefied

639

27.5

19.835.3

HF[P2,3,6]

633

22

20.923.1

23.1

b24.4

c

8.3

8.4

5.3

13.4

4.6

3.28.0

0.37

0.340.36

186


11/20

habitat into adjacent habitat types. Compared to closedforests, heterogeneous landscapes such as the forest-savannamosaic of CNP comprise much more edge habitat(Hennenberg et al. 2005), thus offering more opportunitiesfor species foraging in this situation. When broadly viewedon the plot level, none of the estimates of species richnessapproached the respective values on the landscape scale,indicating that spatial patterns of species aggregation andturnover found on the local scale precluded accurateestimation of species diversity on the landscape scale.

Our results contradict those of Rautenbach et al. (1996),who studied the effects of structural complexity on batspecies richness along a latitudinal transect across Kruger

National Park, South Africa. They found no significantdifferences in diversity patterns between paired plots ingallery forests and woodlands. While their study of batassemblages is one of the few with a largely standardizedsampling protocol, sampling bias has probably seriouslyimpacted their results. First, it is likely that many specieswere missed since neither elevated nets nor harp traps wereused. Second, edge habitat but not the forest itself wassampled with large mist nets employed perpendicular togallery forest, which probably explains the surprisingly lownumber of species in gallery forests adapted to densevegetation (Rautenbach et al. 1996).

To summarize, we propose that the generality of

species diversity-habitat complexity relationships is criti-cally dependent on how much a given group utilizes thehabitat along the vertical axis. While some studies of non-volant mammals revealed links between species diversityand habitat complexity, others did not (August 1983,

Williams et al. 2002, and references therein). The classicstudies by MacArthur (MacArthur and MacArthur 1961,MacArthur 1964), however, demonstrated a strikingrelationship between species diversity of birds and struc-tural complexity, which has been recently corroboratedwith high-resolution remote sensing data (Goetz et al.2007).

Structure of functional groups in relation to habitattype

Earlier studies demonstrated a close link between ecomor-phological and -physiological characters of bats and theirhabitat-specific foraging patterns, which were in turn relatedto the structure of species assemblages (McKenzie and Rolfe1986, Aldridge and Rautenbach 1987, Crome and Richards1988). We expected that the proportional composition offunctional groups would mirror the different habitat typeswhen compared to the composition on the landscape scale.

We predicted that three functional groups foraging in densehabitat (FgNS, AgNS, AaNS) should be overrepresented in

TNP-plots as well as in forest habitat in CNP (FI, GF) whilethe group foraging in open space (AaOS) should beunderrepresented. We further expected the opposite of thispattern in savanna plots of CNP. For the group foraging inedge habitat (AaEG) we anticipated proportional samplingin both CNP and TNP. These predictions were onlypartially confirmed (Table 6). In CNP, functional groupsshowed partly idiosyncratic responses. For example, thefunctional group composition of gallery forests constituted arandom draw compared to the species pool of the landscapescale whereas two functional groups (FgNS and AaOS)showed significant deviations in forest islands in accordancewith our prediction. Two of four functional groups (AaNS

and AaOS) deviated from proportional sampling in thesavanna assemblage of CNP according to our expectationand another group (FgNS) showed a similar trend. Bothforest types of TNP showed the closest agreement with ourpredictions, with two functional groups (FgNS and AaOS)revealing a significant departure and a third (AaNS) evincinga trend. The proportionally lower presence of the AaOSgroup should be treated with caution, however, since thisgroup is notoriously difficult to sample with mist nets andharp traps, especially in forest habitat such as TNP wherethis group was only within reach of our elevated mist nets inlarge canopy gaps.

Table 6. Functional group composition of bat assemblages in relation to habitat type: observed compared to expected species richness(10 000 randomizations: mean and 95% CI in parentheses; significant differences [proportion of observed values 5 or ] randomizations]shown in italics). Functional group classification modified from Schnitzler and Kalko (2001; Appendix): F frugi- and nectarivorous, A animalivorous, gNS gleaning narrow space, aNS aerial narrow space, aEG aerial edge and gap, aOS aerial open space; Landscape:pooled data from plots and opportunistic sampling, SF: swamp forests; HS: hill forests, FI: forest islands, GF: gallery forests, SA: savannas.

Sample Functional group Total

FgNS AgNS AaNS AaEG AaOS

Comoe NP

Landscape 9 9 8 17 14 57FIobs 9 (p00.0008) 4 6 7 2 (p00.0029) 28FIexp 4.4 (27) 4.4 (27) 3.9 (16) 8.3 (512) 6.9 (410)GFobs 8 6 5 11 9 39GFexp 6.2 (48) 6.1 (48) 5.5 (38) 11.6 (815) 9.6 (712)SAobs 8 3 2 (p00.0368) 9 12 (p00.0214) 34SAexp 5.4 (38) 5.4 (38) 4.8 (27) 10.1 (713) 8.3 (511)

Ta NPLandscape 8 7 7 10 8 40SFobs 8 (p00.0443) 4 7 7 2 (p00.0048) 28SFexp 5.6 (38) 4.9 (37) 4.9 (37) 7.0 (49) 5.6 (38)HFobs 8 (p00.0036) 3 6 4 1 (p00.0092) 22HFexp 4.4 (27) 3.9 (26) 3.9 (26) 5.5 (38) 4.4 (27)

187


12/20

We expected that habitat type would lead to a morepronounced shift in proportional composition of functionalgroups. However, since the null model compared observedand expected values based on incidence data, possible shiftsin the relative abundance structure of assemblages remainedundetected. Moreover, the occasional sampling of speciesusually foraging in a different habitat type might haveblurred our analysis especially for the habitat mosaic ofCNP. Most interestingly, the group composition of both

landscape assemblages (CNP and TNP) constituted almostperfect random draws from the regional species pool. Weanticipated this result for CNP, where the habitat mosaicshould lead to proportional sampling when analyzed on thelandscape scale (i.e. pooling group compositions fromdistinct habitat types). This result was surprising forTNP, where we expected a preponderance of functionalgroups adapted to dense vegetation (FgNS, AgNS, AaNS)as well as a lower proportion of species foraging in openhabitat (AaOS), and opens intriguing questions to whichextent species richness on local to landscape scales isgoverned by the availability of suitable habitat (bottomup) or if overlapping distribution ranges determine the

regional species pool in a way leading to proportionalsampling on landscape to local scales (top down). Alter-natively, our classification of functional groups might havebeen too coarse to detect a clear signal when analyzing thepooled communities on the landscape level.

Habitat heterogeneity and partitioning of diversitycomponents

Beta diversity as measured on the landscape scale is afunction of two non-exclusive mechanisms. First, if speciesare characterized by high habitat specificity, increase inhabitat heterogeneity should result in increased beta

diversity. Second, dispersal limitation can lead to higherbeta diversity, particularly in the case of spatially discontin-uous habitat patches (Mouquet and Loreau 2003, Freestoneand Inouye 2006). Kadmon and Allouche (2007) modelledcomplex relationships between habitat heterogeneity andspecies richness depending on area effects as well as dispersaland reproductive rates, and habitat heterogeneity only had amonotonically positive effect on species richness whendispersal rates were high and habitat patches were large.

In the habitat mosaic of CNP, heterogeneity had astrong and positive effect on beta diversity, which wassignificantly higher among habitat types (b1) as well asbetween replicates (b2) than expected (additive partition-

ing). On the contrary, the rather uniform TNP did notshow elevated patterns of beta diversity; i.e. landscaperichness was neither significantly increased by a highturnover of species between habitat types (b1) nor on thesecond level of our analysis (among replicates: b2). SinceCNP had a significantly higher mean plot diversity thanTNP at the local scale (both Sobs and Sest), we asked whetherthe higher landscape diversity of CNP was driven mainly bylocal (alpha) or beta diversity. The additive partitioningapproach revealed that both factors contributed aboutequally to landscape diversity in CNP, but local diversitywas significantly lower and beta diversities were significantlyhigher than expected by the null model. In TNP, the local

scale contributed much more to the landscape scale (59.1%)than in CNP, although local diversity was likewise reducedcompared to expected values, while beta diversities did notdeviate from the null expectations.

Further support for a causal relationship between habitatheterogeneity and landscape richness is shown by the spatialpartitioning of beta diversities on the two hierarchical scalesin CNP. The nested plots representing different habitattypes contributed almost two-thirds to landscape diversity

although plots on this level were spatially much closer toeach other than on the next hierarchical level. In a habitatmosaic such as CNP, community similarity should varyfrom high to low depending on whether similar ordissimilar habitat types are being compared. Beta diversityshould correlate with geographic distance only if dispersalability is low, thus imposing a filter on communitysimilarity with increasing distance (Nekola and White1999, Freestone and Inouye 2006). However, this wasrejected for CNP by the Mantel test, indicating thatdispersal limitations between habitat types were negligibleat the distances covered by our study and due to the highmobility of bats. This matches results by Veech and Crist

(2007), who found no decay of community similarity forbird assemblages within North American ecoregions but apositive relation between habitat heterogeneity and birddiversity at the landscape scale.

Furthermore, we employed a grain size that was smallerthan the patch size of distinct habitat types, therebyallowing for habitat-specific responses in assemblage com-position. Larger grain sizes would have averaged betadiversity over different habitat types (Nekola and White1999), such that changes in species turnover could havebeen detected only over larger distances and lineargradients. The lack of decay in community similarity withgeographical distance in CNP matched the spatial config-uration of the study region. Here, plots were located in

distinct habitat types that were defined by sharp and stableboundaries (Hennenberg et al. 2005, Goetze et al. 2006). Incontrast, structurally defined habitat types in the rainforestregion of TNP lacked pronounced boundaries. Parallel tothe gradual rather than abrupt changes in floristic composi-tion with local topography in TNP (Van Rompaey 1993),turnover in the composition of bat assemblages wasunrelated to structurally defined habitat types. Mantel testsshowed that the significant variation of beta diversity withdistance was the result of very similar assemblage structuresin neighbouring plots (paired sampling design) althoughthey represented different forest types. This pattern brokedown when we used average similarity values for the three

neighbouring plots pairs, indicating that habitat differenceshad a minor influence on assemblage patterns at this spatialscale. Interestingly, community composition of leaf-litteranurans in the very same plots in TNP was explained bygeographic distance rather than by environmental variables(Ernst and Rodel 2005). Here, low mobility of leaf-litteranurans seems to result in assemblages that are predicted bygeographic distance rather than environmental factors.

High mobility or dispersal rates do not necessarily leadto homogenization of assemblages in distinct habitatpatches if the latter are characterized by contrastingenvironments. In CNP, forest and savannas are structurallyvery different, hence species adapted to dense habitats

188


13/20

should have disadvantages when foraging in open savannaswith strongly reduced complexity, while species adapted toopen habitats might be almost completely excluded fromentering dense habitat because of ecomorphologicaland -physiological constraints (Aldridge and Rautenbach1987, Schnitzler and Kalko 2001). In our view, thestructure of bat metacommunities on the landscape scalemight be largely regulated by the filter properties imposedby physical habitat parameters (e.g. vegetation density),

rather than by species interactions within a given habitat.Thus, we suggest differentiating between strict dispersallimitation on the one hand (i.e. organisms with lowmobility, or those that rarely cross unsuitable habitat) andon the other hand high dispersal but low recruitment due tospecific habitat preferences (i.e. organisms that regularlycross or encounter unsuitable habitat without establishmentof permanent populations). We suggest that variation inlocal habitat conditions lead to intraspecific aggregation ofbats species to the extent that these conditions meet theirhabitat requirements. Intraspecific aggregation may reducelocal diversity (He and Legendre 2002, Veech et al. 2003),in agreement with our findings of reduced diversity at the

local scale (plots) both in TNP and CNP, albeit with amuch higher effect in CNP. The latter finding is inaccordance with the pronounced contrasts between savannaand forest habitats in CNP, which should result in higherintraspecific aggregation in relation to habitat type (i.e.within distinct patches), and a stronger reduction of localdiversity when compared to expectations of randomplacement.

Our study is one of few to assess the influence of habitatheterogeneity across spatial scales in natural habitat mosaics;most previous research has been conducted in anthropo-genically fragmented landscapes (Tews et al. 2004). Recentanthropogenic fragmentation might lead to qualitatively

and quantitatively different patterns as the regional speciespool from which local assemblages are recruited might befundamentally different from a natural biome transitionsuch as CNP, where habitat types as well as distributionranges of species interdigitate. In Paraguay, Stevens et al.(2004) documented much higher diversity of bats in aforest-savanna mosaic compared to a fairly uniform forestregion and linked this pattern to the effects of habitatheterogeneity. Similarly, the importance of forest habitats inthe southern part of CNP, which cover a mere 10.7% ofthis region (Hovestadt et al. 1999), is illustrated by the highspecies overlap between CNP and TNP where more thanhalf of the species recorded in TNP also occur in CNP. Inthis context, gallery forests along larger rivers such as theComoe might be critical for linking forest populations inthe rainforest zone with more isolated populations in theforest-savanna mosaic to the north. Habitat heterogeneityalso predicts diversity patterns of non-volant mammals in

Australia (Williams et al. 2002). The latter study demon-strated scale-dependency, where heterogeneity became anexcellent predictor of species richness at larger spatial scales,showing that the effects of heterogeneity depend on theperceived grain of the study organism. The lack ofheterogeneity-diversity relationships reported by Cramerand Willig (2005) might be explained by the microscale atwhich they studied rodents.

Mixture effects of habitat complexity andheterogeneity on species richness across biomes

Since a vast majority of studies evinced a positive relation-ship between habitat heterogeneity and species diversity(reviewed by Tews et al. 2004), we sought to conceptualizeand integrate the effects of complexity and heterogeneityalong a highly simplified gradient of biomes from forests tosavannas and steppes, building on a virtual transect across

West Africa (Fig. 5). As one moves from rainforests in thesouth to steppe in the north, we posit that habitat

complexity decreases monotonically as vegetation strataare increasingly lost with diminishing stature of woodyplants, finally giving way to biomes that are dominated byherbs and low bushes. Along this gradient, we conjecturethat habitat heterogeneity is rather low over a broad climaticrange, specifically, within the forest zone where canopy isbroken mainly by elements such as treefall gaps and watercourses. Once a critical threshold is crossed, increasedseasonality and/or reduced precipitation leads to a biometransition between forests and savannas (Sankaran et al.2005). Within this ecotone, habitat heterogeneity increasessharply until forest elements such as forest islands andgallery forests are completely lost further north. For flying

organisms such as bats and birds utilizing both the verticaland horizontal axis, we hypothesize that local (alpha)diversity is driven mainly by habitat complexity whilebeta diversity results from heterogeneity. If true, landscapediversity would result from the joint effects of complexityand heterogeneity, giving rise to a pronounced peak withinthe biome transition between forests and savannas. Amixture model of habitat complexity and heterogeneityhas been also invoked by Roth (1976) to explain contrastingpatterns of bird diversity from North American grasslandsto forests.

We compared patterns documented by our study for batswith patterns of species richness of birds in CNP and TNP.

Wet - dry gradient

rain forest dry forest biome transition savanna steppe

Total effectComplexityHeterogeneity

Figure 5. Hypothetical model of the contribution of habitatstructure on species richness along a forest-savanna-steppe gradi-ent: habitat complexity decreases along this gradient withdiminishing vegetation height, habitat heterogeneity shows anintermediate peak at the biome transition between forests and

savannas, and the effects of habitat complexity and heterogeneity,in their sum, lead to a diversity peak at the biome transition.

189


14/20

In CNP, 494 bird species have been recorded (Salewski2000). When we subtract Palaearctic migrants, which aremainly occurring in northern CNP, and aquatic species, 373species remain for the southern part of CNP. For TNP, morethan 230 bird species have been recorded (Gartshore et al.1995), and 215 remain after excluding aquatic species andPalaearctic migrants. In contrast to bats and birds, plantdiversity shows the opposite pattern, with 1233 speciesknown from TNP and 720 species from CNP (Poilecot

1991, Hovestadt et al. 1999, Dengueadhe Kolongo et al.2006). Parallel to plant diversity, actual and potential annualevapotranspiration are higher in TNP than in CNP (Tateishiand Ahn 1996: TNP: 1397 mm AET, 1515 mm PET; CNP:1027 mm AET, 1388 mm PET). Estimated annual netprimary production is almost twice as high in TNP as inCNP (Imhoff et al. 2004: TNP: 1177.6 g C m(2, CNP:609.9 g C m(2). Contrary to hypotheses developed on theregional scale (Hawkins et al. 2003), landscape diversity ofbats and birds does not match the marked differences inenergy availability, productivity, precipitation, or floristicdiversity. In conclusion, we suggest that species richness ofbats and birds follows a mixture model integrating habitat

complexity and heterogeneity, with peak richness found inthe forest-savanna biome transition. In agreement with ourstudy, Williams et al. (1999) documented pronounced peaksof both species richness and turnover for sub-Saharan birds inthe transition zone bordering equatorial forests. Moreover,Kark et al. (2007) demonstrated that bird species richnesspeaks along ecoregion boundaries of the New World.Bridging these studies, which employed coarse samplinggrains over large geographical extents, our study supports thenotion that biome transitions harbour significantly increaseddiversity from local to landscape scales.

Biome transitions and ecotones are increasingly seen assignificant centres of evolutionary processes for the main-tenance, and possibly generation, of diversity (Moritz et al.

2000, Spector 2002). Our study lends further support tothe importance of heterogeneous habitats for conservationstrategies aiming at safeguarding ecological and evolution-ary diversification (Kark et al. 2007). As a focal point,keystone structures within heterogeneous landscapesdeserve particular consideration (Tews et al. 2004). In thehabitat mosaic of CNP, these keystone structures constituteforest islands and gallery forests as they allow for thepersistence of forest-adapted species in the surroundingsavanna matrix. However, forest patches are increasinglylost through land use outside the protected area of CNP(Goetze et al. 2006), with potentially far-reaching con-sequences for the maintenance of ecological and evolu-

tionary processes such as gene flow between populationsalong corridors such as gallery forests. Thus, systematicconservation planning should specifically consider biometransitions with an evaluation of their spatial connectivity toneighbouring source areas.

Species richness of Afrotropical bats

In a first attempt to characterize regional species richness ofbats globally, Findley (1993) suggested that the Afrotropicalregion is impoverished compared to the Australasian andNeotropical realms. He further concluded that species

richness of African bats does not peak in equatorial forestsbut rather in the grasslands and savannas of East Africa,where, according to his data, species richness reaches 6070species per 250 000 km2 grid cell compared to 100120

species in the equatorial regions of the Neotropics andsouth-east Asia. Although this pattern has been widely cited(Kingston et al. 2003, Willig et al. 2003, Proches 2005), weargue that species richness of bats has been largely under-estimated in the Afrotropics.

Compared to the few available data, species richness ofbats in CNP and TNP by far exceeds any African sitessurveyed so far on similar spatial scales. Our figures fromCNP even surpass documented bat richness of vast areassuch as Kruger NP, Garamaba NP, or the Ivindo Basin(Table 7). Furthermore, what Findley (1993) called East

African grasslands and savannas in fact comprises theAlbertine Rift, the Eastern Arc Mountains, and the coastalforests of East Africa, all regions distinguished by pro-

nounced habitat heterogeneity and known to harbour bothhigh diversity and levels of endemism (Brooks et al. 2001).Second, East Africa has been historically much betterexplored than most of the Central and West Africanregions, hence data on species richness of Africa bats shouldbe carefully evaluated against sampling artefacts. Finally,species richness in CNP compares well with results fromstudies in the Neotropics and Australasia, refuting a generalimpoverishment of Afrotropical bat assemblages. Futurecomparisons should be made with great caution to accountfor confounding effects such as sampling methods, sam-pling effort, and area effects.

Conclusions and future perspectives

We believe that the mixture model of habitat heterogeneityand complexity on species richness is likely to have generalapplicability for African bats on local to landscape scales,and possibly also for other taxonomic groups such as birds.It will be interesting to see whether this pattern holds truefor bats in other tropical regions when studied on similarspatial scales. Idiosyncratic differences among continentswill be particularly exciting to analyze both with respect tothe historic legacies of these regions and potential evolu-tionary constraints such as niche conservatism or divergent

Table 7. Documented (Sobs) and estimated (Sest) species richness ofAfrotropical bats in relation to area (CNP and TNP #: estimatedfrom samples, see Table 2; : estimated from nearby records withinthe same biome, see Supplementary material; other areas 1: Rautenbach et al. 1996, 2: Verschuren 1957, 3: Brosset 1966).

Study region Sobs Sest Area [km2]

Comoe NP 50 656 6063# 4557 170

6571

entire park(11500)Ta NP 40 4144# 6

4450 entire park(4500)

Kruger NP, South Africa1 41 ca 20 000Garamba NP, D.R. Congo2 ca 40 ca 5000Ivindo Basin, Gabon3 34 ca 2500

190


15/20

radiations (Gavrilets and Vose 2005, Wiens and Graham2005, Stevens 2006).

Acknowledgements We greatly appreciate the contribution ofStefan Pettersson, Njikoha Ebigbo and Katja Soer to the data setanalyzed here. We thank Kouakou Kouame, Koffi Kouadio andKouadio Kouakou (all CNP) and Georges Gbamlin (TNP) fortheir dedicated assistance in the field. Vital logistical support wasprovided by K. Eduard Linsenmair, Frauke Fischer and theemployees of the Projet Biodiversite at CNP, and by theCentre de Recherche en Ecologie, the Projet Autonome pour laConservation du Parc National de Ta , and the Ta MonkeyProject at TNP. Research permits to work in CNP and TNPwere kindly granted by the Ministere de lAgriculture et desRessources Animales and the Ministere de lEnseignementSuperieur et de la Recherche Scientifique, Republique de CotedIvoire. Dieter Kock, Senckenberg Museum Frankfurt, offeredinvaluable help with species identification. Thomas Crist, LouJost, Martin Pfeiffer, Erica Sampaio and Joseph Veech helped invarious ways with the development and interpretation of ouranalyses and in shaping our ideas. We appreciate general supportby Hans-Ulrich Schnitzler and Mark-Oliver Rodel. This is a

contribution of the BIOTA program, funded by the GermanFederal Ministry of Education and Research (BMBF, project01LC0017, 01LC0411 and 01LC0617E1). We acknowledgeadditional funding by the Landesgraduiertenforderung BW andthe German Academic Exchange Service (DAAD). Douglas Kelt,Egbert Leigh, Christoph Meyer and three reviewers helped toimprove previous versions of the manuscript.

References

Aldridge, H. D. J. N. and Rautenbach, I. L. 1987. Morphology,echolocation and resource partitioning in insectivorous bats.J. Anim. Ecol. 56: 763778.

August, P. V. 1983. The role of habitat complexity andheterogeneity in structuring tropical mammal communities. Ecology 64: 14951507.

Bohning-Gaese, K. 1997. Determinants of avian species richness atdifferent spatial scales. J. Biogeogr. 24: 4960.

Brooks, T. M. et al. 2001. Toward a blueprint for conservation inAfrica. BioScience 51: 613624.

Brose, U. and Martinez, N. D. 2004. Estimating the richness ofspecies with variable mobility. Oikos 105: 292300.

Brosset, A. 1966. Les Chiropteres du Haut-Ivindo (Gabon). Biol. Gabon 2: 4786.

Buzas, M. A. and Hayek, L.-A. C. 1996. Biodiversity resolution:an integrated approach. Biodivers. Lett. 3: 4043.

Ceballos, G. and Ehrlich, P. R. 2006. Global mammal distribu-tions, biodiversity hotspots, and conservation. Proc. Nat.

Acad. Sci. USA 103: 1937419379.Chao, A. and Shen, T.-J. 2006. SPADE: species prediction and

diversity estimation. Version 3.1. Program and users guide..

Colwell, R. K. 2005. EstimateS: statistical estimation of speciesrichness and shared species from samples. Version 7.5.Program and users guide. .

Colwell, R. K. et al. 2004. Interpolating, extrapolating, andcomparing incidence-based species accumulation curves. Ecology 85: 27172727.

Cornell, H. V. and Lawton, J. H. 1992. Species interactions, localand regional processes, and limits to the richness of ecologicalcommunities: a theoretical perspective. J. Anim. Ecol. 61:112.

Cramer, M. J. and Willig, M. R. 2005. Habitat heterogeneity,species diversity and null models. Oikos 108: 209218.

Crist, T. O. and Veech, J. A. 2006. Additive partitioning ofrarefaction curves and speciesarea relationships: unifying a-,b- and g-diversity with sample size and habitat area. Ecol.Lett. 9: 923932.

Crist, T. O. et al. 2003. Partitioning species diversity acrosslandscapes and regions: a hierarchical analysis of a, b, and gdiversity. Am. Nat. 162: 734743.

Crome, F. H. J. and Richards, G. C. 1988. Bats and gaps:

microchiropteran community structure in a Queensland rainforest. Ecology 69: 19601969.

Davies, R. G. et al. 2007. Topography, energy and the globaldistribution of bird species richness. Proc. R. Soc. B 274:11891197.

Dengueadhe Kolongo, T. S. et al. 2006. Plant species diversity inthe southern part of the Ta National Park (Cote dIvoire). Biodivers. Conserv. 15: 21232142.

Ernst, R. and Rodel, M.-O. 2005. Anthropogenically inducedchanges of predictability in tropical anuran assemblages. Ecology 86: 31113118.

Fenton, M. B. and Bell, G. P. 1981. Recognition of species ofinsectivorous bats by their echolocation calls. J. Mammal.62: 233243.

Fenton, M. B. and Griffin, D. R. 1997. High-altitude pursuit of

insects by echolocating bats. J. Mammal. 78: 247250.Findley, J. S. 1993. Bats a community perspective. Cambridge

Univ. Press.Freestone, A. L. and Inouye, B. D. 2006. Dispersal limitation and

environmental heterogeneity shape scale-dependent diversitypatterns in plant communities. Ecology 87: 24252432.

Gartshore, M. E. et al. 1995. Forest birds in Cote dIvoire. Asurvey of Ta National Park and forestry plantations, 19891991. Birdlife Int. Study Rep. 58: i-v, 181.

Gavrilets, S. and Vose, A. 2005. Dynamic patterns of adaptiveradiation. Proc. Nat. Acad. Sci. USA 102: 1804018045.

Goetz, S. et al. 2007. Laser remote sensing of canopy habitatheterogeneity as a predictor of bird species richness in aneastern temperate forest, USA. Remote Sens. Environ. 108:

254

263.Goetze, D. et al. 2006. Dynamics of forest-savanna mosaics innorth-eastern Ivory Coast from 1954 to 2002. J. Biogeogr.33: 653664.

Gotelli, N. J. and Colwell, R. K. 2001. Quantifying biodiversity:procedures and pitfalls in the measurement and comparison ofspecies richness. Ecol. Lett. 4: 379391.

Hawkins, B. A. et al. 2003. Energy, water, and broad-scalegeographic patterns of species richness. Ecology 84: 31053117.

He, F. and Legendre, P. 2002. Species diversity patterns derivedfrom speciesarea models. Ecology 83: 11851198.

Hennenberg, K. J. et al. 2005. Border and ecotone detection byvegetation composition along forest-savanna transects in IvoryCoast. J. Veg. Sci. 16: 301310.

Hill, M. O. 1973. Diversity and evenness: a unifying notation andits consequences. Ecology 54: 427432.Hillebrand, H. 2004. On the generality of the latitudinal diversity

gradient. Am. Nat. 163: 192211.Hovestadt, T. et al. 1999. Seed dispersal mechanisms and the

vegetation of forest islands in a West African forest-savannamosaic (Comoe National Park, Ivory Coast). Plant Ecol.144: 125.

Hurlbert, A. H. and White, E. P. 2005. Disparity between rangemap- and survey-based analyses of species richness: patterns,processes and implications. Ecol. Lett. 8: 319327.

Hurlbert, A. H. and Jetz, W. 2007. Species richness, hotspots, andthe scale dependence of range maps in ecology and conserva-tion. Proc. Nat. Acad. Sci. USA 104: 1338413389.

191
http://chao.stat.nthu.edu.tw/http://purl.oclc.org/estimateshttp://purl.oclc.org/estimateshttp://chao.stat.nthu.edu.tw/


16/20

Imhoff, M. L. et al. 2004. Global patterns in human consumptionof net primary production. Nature 429: 870873.

Jost, L. 2006. Entropy and diversity. Oikos 113: 363375.Kadmon, R. and Allouche, O. 2007. Integrating the effects of area,

isolation, and habitat heterogeneity on species diversity: aunification of island biogeography and niche theory. Am.Nat. 170: 443454.

Kark, S. et al. 2007. The role of transitional areas as avianbiodiversity centres. Global Ecol. Biogeogr. 16: 187196.

Kerr, J. T. and Packer, L. 1997. Habitat heterogeneity as a

determinant of mammal species richness in high-energyregions. Nature 385: 252254.

Kerr, J. T. et al. 2001. Remotely sensed habitat diversity predictsbutterfly species richness and community similarity in Canada.

Proc. Nat. Acad. Sci. USA 98: 1136511370.Kingston, T. et al. 2003. Species richness in an insectivorous bat

assemblage from Malaysia. J. Trop. Ecol. 19: 6779.Lande, R. 1996. Statistics and partitioning of species diversity, and

similarity among multiple communities. Oikos 76: 513.Loreau, M. 2000. Are communities saturated? On the relationship

between a, b and g diversity. Ecol. Lett. 3: 7376.Lyons, S. K. and Willig, M. R. 1999. A hemispheric assessment of

scale dependence in latitudinal gradients of species richness. Ecology 80: 24832491.

Lyons, S. K. and Willig, M. R. 2002. Species richness, latitude,and scale-sensitivity. Ecology 83: 4758.MacArthur, R. H. 1964. Environmental factors affecting bird

species diversity. Am. Nat. 98: 387397.MacArthur, R. H. and MacArthur, J. W. 1961. On bird species

diversity. Ecology 42: 594598.McCune, B. and Mefford, M. J. 2006. PC-Ord for Windows.

Multivariate analysis of ecological data. Version 5.07. MjMSoftware, Gleneden Beach, OR, USA.

McKenzie, N. L. and Rolfe, J. K. 1986. Structure of bat guilds inthe Kimberley mangroves, Australia. J. Anim. Ecol. 55: 401420.

Moritz, C. et al. 2000. Diversification of rainforest faunas: anintegrated molecular approach. Annu. Rev. Ecol. Syst. 31:533563.

Mouquet, N. and Loreau, M. 2003. Community patterns insource-sink metacommunities. Am. Nat. 162: 544557.

Nekola, J. C. and White, P. S. 1999. The distance decay ofsimilarity in biogeography and ecology. J. Biogeogr. 26:867878.

Neuweiler, G. 1989. Foraging ecology and audition in echolocat-ing bats. Trends Ecol. Evol. 4: 160166.

Norberg, U. M. and Rayner, J. M. V. 1987. Ecologicalmorphology and flight in bats (Mammalia; Chiroptera):wing adaptations, flight performance, foraging strategy andecholocation. Phil. Trans. R. Soc. B 316: 335427.

Olson, D. M. et al. 2001. Terrestrial ecoregions of the world: anew map of life on Earth. BioScience 51: 933938.

Orme, C. D. L. et al. 2006. Global patterns of geographic range

size in birds.

PLoS Biol. 4: e208.Poilecot, P. 1991. Un ecosysteme de savane soudanienne: Le ParcNational de la Comoe (Cote dIvoire). UNESCO.

Proches, S. 2005. The worlds biogeographical regions: clusteranalyses based on bat distributions. J. Biogeogr. 32: 607614.

Rahbek, C. 2005. The role of spatial scale and the perception oflarge-scale species-richness patterns. Ecol. Lett. 8: 224239.

Rahbek, C. and Graves, G. R. 2001. Multiscale assessment ofpatterns of avian species richness. Proc. Nat. Acad. Sci. USA98: 45344539.

Rautenbach, I. L. et al. 1996. Bats in riverine forests andwoodlands: a latitudinal transect in southern Africa. Can.J. Zool. 74: 312322.

Resampling Stats 2006. Resampling Stats for Excel. Version 3.2.Program and users guide. .

Ricklefs, R. E. 1987. Community diversity: relative roles of localand regional processes. Science 235: 167171.

Ricklefs, R. E. 2004. A comprehensive framework for globalpatterns in biodiversity. Ecol. Lett. 7: 115.

Rosenzweig, M. L. 1995. Species diversity in space and time. Cambridge Univ. Press.

Roth, R. R. 1976. Spatial heterogeneity and bird species diversity. Ecology 57: 773782.

Ruggiero, A. and Kitzberger, T. 2004. Environmental correlates ofmammal species richness in South America: effects of spatialstructure, taxonomy and geographic range. Ecography 27:401417.

Safi, K. and Dechmann, D. K. N. 2005. Adaptation of brainregions to habitat complexity: a comparative analysis in bats(Chiroptera). Proc. R. Soc. B 272: 179186.

Salewski, V. 2000. The birds of Comoe National Park, IvoryCoast. Malimbus 22: 5576.

Sankaran, M. et al. 2005. Determinants of woody cover in Africansavannas. Nature 438: 846849.

Scheiner, S. M. et al. 2000. Species richness, speciesarea curvesand Simpsons paradox. Evol. Ecol. Res. 2: 791802.

Schnitzler, H.-U. and Kalko, E. K. V. 2001. Echolocation byinsect-eating bats. BioScience 51: 557569.

Schoeman, M. C. and Jacobs, D. S. 2003. Support for theallotonic frequency hypothesis in an insectivorous bat com-munity. Oecologia 134: 154162.

Schoeman, M. C. and Jacobs, D. S. 2008. The relative influence ofcompetition and prey defenses on the phenotypic structure ofinsectivorous bat ensembles in southern Africa. PLoS One 3:e3715.

Shmida, A. and Wilson, M. V. 1985. Biological determinants ofspecies diversity. J. Biogeogr. 12: 120.

Spector, S. 2002. Biogeographic crossroads as priority areas forbiodiversity conservation. Conserv. Biol. 16: 14801487.

Stevens, R. D. 2006. Historical processes enhance patterns ofdiversity along latitudinal gradients. Proc. R. Soc. B 273:22832289.

Stevens, R. D. et al. 2004. Comparative community ecology ofbats from eastern Paraguay: taxonomic, ecological, andbiogeographic perspectives. J. Mammal. 85: 698707.

Summerville, K. S. et al. 2003. Spatial variation in species diversityand composition of forest Lepidoptera in eastern deciduousforests of North America. Conserv. Biol. 17: 10451057.

Tateishi, R. and Ahn, C. H. 1996. Mapping evapotranspirationand water balance for global land surfaces. ISPRS J.Photogramm. 51: 209215.

Taylor, P. J. 1999. Echolocation calls of twenty southern Africanbat species. S. Afr. J. Sci. 34: 114124.

Taylor, P. J. et al. 2005. Intraspecific variation in the calls of someAfrican bats (Order Chiroptera). Durban Mus. Novit. 30:2437.

Tews, J. et al. 2004. Animal species diversity driven by habitat

heterogeneity/diversity: the importance of keystone structures.J. Biogeogr. 31: 7992.

Tuomisto, H. and Ruokolainen, K. 2006. Analyzing or explainingbeta diversity? Understanding the targets of different methodsof analysis. Ecology 87: 26972708.

Van Rensburg, B. J. et al. 2002. Species richness, environmentalcorrelates, and spatial scale: a test using South African birds.Am. Nat. 159: 566577.

Van Rompaey, R. S. A. R. 1993. Forest gradients in West Africa: aspatial analysis. Ph.D. thesis, Wageningen Agricultural Univ.

Veech, J. A. and Crist, T. O. 2007. Habitat and climateheterogeneity

Documents

Biome transitions as centres of diversity: Habitat heterogeneity and diversity patterns of West African bat assemblages across spatial scales