This article was downloaded by: [UMA University of Malaga] On: 06 September 2011, At: 01:15 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Ostrich Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tost20 Biogeographical zonation of African hornbills and their biotic and geographic characterisations A Román Muñoz, Raimundo Real, Jesús Olivero, Ana L Márquez, José C Guerrero, Silvia B Bárcena & J Mario Vargas Available online: 12 Nov 2009 To cite this article: A Román Muñoz, Raimundo Real, Jesús Olivero, Ana L Márquez, José C Guerrero, Silvia B Bárcena & J Mario Vargas (2003): Biogeographical zonation of African hornbills and their biotic and geographic characterisations, Ostrich, 74:1-2, 39-47 To link to this article: http://dx.doi.org/10.2989/00306520309485368 PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan, sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
This article was downloaded by [UMA University of Malaga]On 06 September 2011 At 0115Publisher Taylor amp FrancisInforma Ltd Registered in England and Wales Registered Number 1072954 Registered office MortimerHouse 37-41 Mortimer Street London W1T 3JH UK
OstrichPublication details including instructions for authors and subscription informationhttpwwwtandfonlinecomloitost20
Biogeographical zonation of African hornbills andtheir biotic and geographic characterisationsA Romaacuten Muntildeoz Raimundo Real Jesuacutes Olivero Ana L Maacuterquez Joseacute C Guerrero SilviaB Baacutercena amp J Mario Vargas
Available online 12 Nov 2009
To cite this article A Romaacuten Muntildeoz Raimundo Real Jesuacutes Olivero Ana L Maacuterquez Joseacute C Guerrero Silvia B Baacutercena ampJ Mario Vargas (2003) Biogeographical zonation of African hornbills and their biotic and geographic characterisationsOstrich 741-2 39-47
To link to this article httpdxdoiorg10298900306520309485368
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use httpwwwtandfonlinecompageterms-and-conditions
This article may be used for research teaching and private study purposes Any substantial or systematicreproduction re-distribution re-selling loan sub-licensing systematic supply or distribution in any formto anyone is expressly forbidden
The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date The accuracy of any instructions formulae and drug dosesshould be independently verified with primary sources The publisher shall not be liable for any lossactions claims proceedings demand or costs or damages whatsoever or howsoever caused arisingdirectly or indirectly in connection with or arising out of the use of this material
Ostrich 2003 74(1amp2) 39ndash47Printed in South Africa mdash All rights reserved
Copyright copy NISC Pty Ltd
OSTRICHISSN 0030ndash6525
Biogeographical zonation of African hornbills and their biotic andgeographic characterisations
A Romaacuten Muntildeoz Raimundo Real Jesuacutes Olivero Ana L Maacuterquez Joseacute C Guerrero Silvia B Baacutercena
and J Mario VargasDepartamento de Biologiacutea Animal Facultad de Ciencias Universidad de Maacutelaga E-29071 Maacutelaga Spain
Corresponding author e-mail romanumaes
We studied the geographic ranges of the 23 African hornbill species using countries as distribution units to search for eithergroups of similar distributions (chorotypes) or gradual replacement of species as two different types of distribution patternsWe used a probabilistic classification method to distinguish between these two types of biogeographic patterns Then weanalysed the influence of climate area surface covered by different types of habitat and human disturbance in each coun-try as well as biotic features of the species involving the habitat preference feeding requirements and habits body size sex-ual dimorphism and metabolic parameters on the biogeographic patterns detected We found that 20 species were signifi-cantly classified into three chorotypes The western-central chorotype comprises nine species characterised by their prefer-ence for forest habitat and their low sexual body mass dimorphism These species occur in 25 countries characterised bytheir low range of temperatures The eastern chorotype includes three species characterised by their preference for feedingon the ground and greater dimorphism in mass between the sexes Seven countries characterised by their high values ofsavanna and grassland surface area constitute the geographic element of this chorotype The southern chorotype involvesthe distribution of eight species characterised by their preference for montane and coastal forests or woodlands Eighteencountries fall into this chorotype and are characterised by their low values of temperature in the coldest month and greatersurface area of desert scrub and desert forest The other three species replace each other mainly following a gradual patternwithin the study area (A Spanish translation of the abstract is provided on p 47)
Hornbills are a group of species that share many featuresfrom the taxonomical and the morphological points of viewThe most obvious characteristics are the long and decurvedbill and the projecting casque unique among avian familiesfeatures that assign them to the order Bucerotiformes(Sibley and Monroe 1990) They occupy an extensive geo-graphical range and from a biogeographic point of viewhave traditionally been considered as exhibiting merely twodisjoint distributions in sub-Saharan Africa and in south-eastAsia respectively (Kemp 1995 2001) The African speciesoccupy a wider range of habitats than the Asian speciesoccurring in a variety of environments from rainforest to aridsavanna The ecological complexity of the African hornbillspecies may result in complex biogeographical trends whicharise from the interactions of the species with their environ-ment and between themselves Because of this complexityit is difficult to integrate the present knowledge of the indi-vidual ecological properties of African hornbill species tounderstand their biogeographical trends It is preferable tofollow a synthetic approach that might account for someemergent properties of the distribution of hornbills and couldbe useful to understand the environmental and biotic influ-ences on the distribution of African hornbills as a whole
The geographical relationship among the distributionpatterns of several species may result in a continuum gradi-ent that is a gradual substitution of species in space or ina discrete pattern that is a set of similar distributions spa-
tially differentiated from other sets of similar distributions(see Mayr 1965 Austin and Smith 1989 Hengeveld 1990)These two kinds of patterns could coexist for differentgroups of species and it is necessary to test which type ofpattern is followed by each species
Baroni-Urbani et al (1978) called a group of similar dis-tributions that can be operatively distinguished from theother species ranges a lsquochorotypersquo Chorotypes have twodifferent components the geographic element which is thewhole area covered by any distribution of the chorotype andthe biotic element (sensu Birks 1987) which is the group ofspecies whose distributions belong to the same chorotypeIn this way chorotype and biotic element are related but aredifferent concepts that refer to discrete biogeographical pat-terns When chorotypes cannot be operatively obtained forsome species then they may be assumed to follow a con-tinuous pattern
Climate (specifically the availability of water and energy)is considered by many authors to be a major macro-envi-ronmental factor determining large scale biogeographicalpatterns (Hengeveld and Hogeweg 1979 Guillet and Crowe1986 Wiens 1989 Saeligtersdal and Birks 1993) For exam-ple the availability of energy has been found to explain abiogeographical regionalisation for waterbirds in Africa(Guillet and Crowe 1985) and the geographical trends in birdspecies richness in several areas (Rabinovich and Rapoport1975 Wright 1983 Turner et al 1988) whereas the avail-
Introduction
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas40
ability of water has also been found to explain the biogeo-graphical patterns of waterbirds in Africa (Guillet and Crowe1985 1986) and in Australia (Kikkawa and Pearse 1969Whitehead et al 1992) The types of habitat availablehuman influences the different biotic characteristics of thespecies and their ecological plasticity can also play a role inthe configuration of their actual distributions Only Croweand Kemp (1988) studied broad scale distribution patterns ofAfrican hornbill species although they did not search forchorotypes nor related the biogeographical patterns tomacroenvironmental characteristics
In this paper we analyse the chorotypes of African horn-bill species and characterise their distribution patternsaccording to the biotic features of the species and the cli-mate habitat characteristics and human influence on theirgeographic elements
Methods
The species and the study areaWe analysed the distribution of the hornbills in theAfrotropical region We considered the four African generawhich contain 23 species (Appendix 1) We used countriesas operative geographic units (OGUs) because hornbills areconspicuous birds well studied at this level of resolution sothat it is not likely that missing data affect the analysesInformation on the species occurrences in the OGUs weretaken from Kemp (1995) and Kemp (2001)
Probabilistic classification analysis to recognisechorotypesReal et al (1992a) applied the approach of McCoy et al(1986) to classification to develop a probabilistic procedurefor recognising chorotypes Following this approach weobtained a matrix of geographical similarities between thedistributions of each pair of species (a and b) using Baroni-Urbani and Buserrsquos (1976) index
where A is the number of countries where the species a ispresent B is the number of countries where the species b ispresent C is the number of countries where the two speciesa and b occur together and D is the number of countriesfrom which the two species a and b are absent This coeffi-cient takes into account shared absences that is countriesoutside the distribution area of both species where otherAfrican hornbills are present and so the similarities are con-sidered in the context of the whole study area and not onlyin the context of the subset of countries that constitute thetwo distributions compared (Real et al 1992b) Conjointabsences are not likely the result of uneven sampling whenusing countries as OGUs However this index gives moreimportance to shared presences and the possibility that twodistributions are considered similar only because of theirshared absences is avoided by multiplying shared absencesby shared occurrences (Baroni-Urbani and Buser 1976)
The grouping of hornbill distributions was made usingthe agglomerative method of classification UPGMA(Unweighted Pair-Group Method using Arithmetic Averages)
(Sneath and Sokal 1973) We prefered a group-averagemethod over a centroid technique because the latter do notyield monotonic results (Lance and Williams 1967 Sneathand Sokal 1973) The result of the classification was repre-sented as a dendrogram
We used the table of critical values in Baroni-Urbani andBuser (1976) to obtain values that are significantly more sim-ilar (+) or significantly more dissimilar (-) than would beexpected at random (Real et al 1992a Maacuterquez et al 1997)
We considered chorotypes to be those clusters that bestcombined the following characteristics a high proportion ofsignificant similarities (+) within the cluster a low proportionof significant dissimilarities (-) within the cluster and a lowproportion of significant similarities (+) between the distribu-tions of the cluster and the distributions of its most similarcluster The degree to which a distribution cluster A com-bines these conditions is provided by the parameterDW(AxA) (McCoy et al 1986) For the mathematical expan-sion of this parameter see Olivero et al (1998) In this waywe computed the DW(AxA) values for every branch of thedendrogram A cluster was considered a chorotype if a)DW(AxA) = 07071 that is the maximum value possible orb) DW(AxA) was positive higher than those of the otherclusters including the distributions involved and a G-test ofindependence (Sokal and Rohlf 1981) which yielded theparameter GW(AxA) showed that the proportion of the lsquo+rsquosigns within the cluster was significantly higher thanbetween it and its most similar cluster The distributions thatdid not fulfil these conditions were considered ungroupedand so they were considered to follow a continuous patternof gradual substitutions in space
The definition of a cluster as chorotype implies that thereis a significant segregation between it and its most similarcluster We also tested the type of segregation (strong orweak) between chorotypes by computing the parameter DS(see McCoy et al 1986) which increases when the propor-tion of significant dissimilarities (lsquo-rsquo signs) is higher betweenthe chorotypes and lower within each of them and the pro-portion of significant similarities is lower between thechorotypes The segregation between chorotypes wasstrong if a) DS = 07071 or b) DS was positive and a G testof independence which yielded the parameter GS showedthat the proportion of lsquo-rsquo signs was significantly higherbetween the chorotypes than within them else the segrega-tion was weak The parameters GS and GW(AxA) follow theChi-square distribution with one degree of freedom
Environmental and biotic characterisation of thechorotypesWe used 27 variables related to climate type of habitat andhuman influences to characterise the geographic element ofeach chorotype (Table 1) and 26 variables related to thehabitat preferences feeding guilds and habits body sizesexual dimorphism and metabolic parameters of the speciesto characterise the corresponding biotic element (Table 2)
We made these characterisations using a forward step-wise logistic regression (Hosmer and Lemeshow 1989) toobtain the variables that significantly increased the probabil-ity of any country or species to be part of the chorotypeaccording to the following equation
B = CxD ( ) + C
CxD ( ) + A + B - C
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 41
p = ey 1 + ey
where p is the probability e is the basis of the Napierian log-arithms and y is a linear combination of the variables (iethe logit equation) The statistical significance of the modelwas evaluated by means of a Chi-squared test and theparameters in the logistic function were estimated by maxi-mum likelihood The significance of variables within themodel was tested using the log likelihood ratio (LR) criterionie the significance of the change in log likelihood when thevariable is removed from the model
Results
The values of the Baroni-Urbani and Buserrsquos index betweenthe distributions of African hornbills are shown in Table 3The matrix of significant similarities between these distribu-tions is shown in Table 4
We identified three chorotypes for the African hornbills(Table 5 Figure 1) involving the distributions of twentyspecies whereas the distributions of the three species leftfollowed a continuous pattern of spatial substitution
Chorotype 1 The biotic element of this chorotype con-sists of nine species (Appendix 1) whose distributions occu-py central-western Africa (Figure 2) Both the preference forforest habitats and the smaller difference in sexual bodymass dimorphism characterise the species belonging to thechorotype according to the following logit equation
The geographic element on the other hand is charac-terised by a low annual temperature range according to thefollowing logit equation
y = 59347 - 08685 (TR)Chorotype 2 The biotic element is constituted by three
species (Appendix 1) all of them belonging to the genusTockus and distributed in eastern Africa (lsquohornrsquo of Africa)(Figure 3) The species of this chorotype have in common ahigh sexual body mass dimorphism
y = -1549 + 012 (Sexual body mass dimorph)Geographically chorotype 2 is dominated by large
expanses of savanna and grasslandy = -27816 + 00001 (SavGrass)Chorotype 3 The biotic element of this chorotype com-
prises eight species (Appendix 1) mainly distributed insouthern Africa (Figure 4) and is characterised by the habi-tat preference for montane and coastal forest and woodland
y = 86779 ndash 855 (Sexual body mass dimorph) + -9479 (Habitat savanna) +622 (Habitat montane and coastal forest) +2355 (Habitat woodland) +6501 (Habitat forest)
y = 395 +
-424 (Habitat savanna) +525 (Habitat montane and coastal forest) +525 (Habitat woodland) -626 (Habitat forest)
Table 1 Variables used to characterise the geographic element ofeach chorotype and their sources
Mean annual precipitation (mm) 1 PMean annual temperature (degC) 1 TColdest month mean temperature (degC) 1 CTWarmest month mean temperature (degC) 1 WTAnnual temperature range (degC) 1 TRTotal land area (ha) 2 AreaTropical land area (ha) 2 TropArid land area (ha) 2 AridSemi-arid land area (ha) 2 SAridHumid land area (ha) 2 HumidTotal forest area (ha) 2 ForestDry forest area (ha) 2 DryFMoist forest area (ha) 2 MoistFMangroves area (ha) 2 MangroveDesertScrub area (ha) 2 DesertScrubWetlandMarsh area (ha) 2 WetlandSavannahGrassland area (ha) 2 SavGrassDesert forest area (ha) 2 DesertFVery dry forest area (ha) 2 VeryDryFDry deciduous forest area (ha) 2 DryDFHill and montane forest area (ha) 2 HillMontFMoist deciduous forest area (ha) 2 MoistFRain forest area (ha) 2 RainFCropland area (ha) 2 CroplandAverage annual deforestation () 2 DeforestAnnual loss of SavannahGrassland () 2 SGLossPopulation density (Person km-2) 2 PopDensGrowth domestic product per capita 3 Gdp
Sources 1 FAO (1984) 2 World Resources Institute (1994) 3 Cordellierand Didiot (1999)
Table 2 Variables used to characterise the biotic element of eachchorotype
Habitat preference HFeeding preference FeedCarnivorous diet CFrugivorous diet FPreferential feeding location FLArboreal feeding location ArbTerrestrial feeding location TerBody mass of males BMassMBody mass of females BMassFBill length of males BillLCasque volume of males CasqueVPrimary area of males PAreaSecondary area of males SAreaWing size of males WingMWing size of females WingFTail size of males TailMTail size of females TailFTarsus length of males TarsusBasal metabolic rate BMRFlight metabolism FMBasal metabolic rateFlight metabolism FMBMR Linear wing loading LWLAspect ratio ARSexual body mass dimorphism SBMDSexual wing length dimorphism SWLDSexual bill length dimorphism SBLD
Source Kemp (1995)
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas42Ta
ble
3V
alu
es o
f th
e s
imila
rity
ind
ex o
f B
aro
ni-U
rba
ni a
nd B
use
r b
etw
ee
n th
e d
istr
ibut
ion
s of
th
e 2
3 A
fric
an
ho
rnb
ill s
peci
es
Sim
ilari
ty v
alu
es
bet
wee
n sp
ecie
s gr
oup
ed
into
a c
ho
roty
pe
are
en
clos
ed in
a b
ox
Spe
cies
num
ber
s a
s in
Ap
pend
ix 1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
231
00
00
60
60
667
069
00
69
00
64
50
605
06
250
594
05
940
65
40
568
016
20
33
90
15
40
204
02
700
000
00
000
00
00
446
044
60
397
211
00
00
854
043
00
43
00
30
50
356
03
310
305
03
050
50
80
349
051
80
59
60
30
80
404
04
040
273
03
730
15
20
259
030
80
273
221
000
050
40
50
40
36
80
377
03
960
368
03
680
47
10
439
047
00
55
60
33
20
441
04
920
347
04
590
20
40
332
038
60
347
11
000
10
00
08
71
08
750
899
08
710
871
07
25
075
30
172
02
50
02
50
02
140
138
01
720
146
02
03
025
00
250
017
22
10
00
08
71
08
750
899
08
710
871
07
25
075
30
172
02
50
02
50
02
140
138
01
720
146
02
03
025
00
250
017
27
10
00
09
460
973
09
440
944
07
85
063
60
226
03
18
03
20
02
760
182
02
270
192
02
69
032
00
320
022
73
10
000
974
09
460
946
07
91
065
10
270
03
47
02
90
02
500
163
02
030
172
02
40
029
00
290
020
34
10
000
973
09
730
76
10
674
021
40
30
40
30
50
263
01
720
215
01
820
25
40
305
030
50
215
51
000
09
440
73
00
698
022
60
31
80
32
00
276
01
820
227
01
920
26
90
320
032
00
227
61
000
07
30
069
80
226
02
57
03
20
02
760
000
02
270
192
02
69
021
50
215
022
78
10
00
059
20
454
05
13
03
74
03
970
326
02
030
172
02
40
037
40
451
039
19
100
00
000
00
00
00
00
00
000
000
00
000
000
00
00
000
00
000
000
013
100
00
85
30
73
00
861
07
190
678
08
210
46
30
254
037
30
393
141
00
00
64
60
774
07
740
515
06
660
39
30
571
064
60
595
151
00
00
869
06
810
588
07
160
53
20
000
039
30
422
161
000
08
430
610
07
950
42
20
323
046
30
490
171
000
04
900
709
00
00
057
80
681
061
018
10
000
856
07
78
000
00
000
000
019
10
000
62
80
000
000
00
000
201
00
00
000
000
00
000
121
000
090
00
839
101
000
094
811
100
0
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 43Ta
ble
4 M
atr
ix o
f si
gni
fican
t si
mila
ritie
s be
twe
en t
he
dis
trib
utio
ns
of
the
23
Afr
ica
n h
orn
bill
spe
cies
S
ign
ifica
nt
sim
ilari
ties
betw
een
sp
eci
es g
rou
ped
into
a c
hor
oty
pe a
re e
nclo
sed
in a
box
S
peci
es n
umb
ers
as
in A
ppe
ndix
1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
230
++
++
00
00
+0
--
--
--
--
00
021
+0
0-
0-
--
0-
00
-0
0-
0-
--
-22
00
00
00
00
00
0-
00
-0
--
0-
1+
++
++
++
+-
--
--
--
--
--
2+
++
++
++
--
--
--
--
--
-7
++
++
++
--
--
--
--
--
-3
++
++
+-
--
--
--
--
--
4+
++
+-
--
--
--
--
--
5+
++
--
--
--
--
--
-6
++
--
--
--
--
--
-8
00
00
0-
--
-0
00
9-
--
--
--
--
--
13+
++
++
+0
-0
014
++
+0
+0
0+
015
++
0+
0-
00
16+
0+
0-
00
170
+-
0+
018
++
--
-19
0-
--
20-
--
12+
+10
+11 +
S
imila
rity
sig
nifi
cant
ly h
igh
er t
han
ran
dom
exp
ecta
tion
-
Sim
ilarit
y si
gn
ifica
ntly
low
er t
han
ran
dom
exp
ecta
tion
0
S
imila
rity
with
out
sig
nifi
can
t d
iffer
ence
with
the
ra
ndom
exp
ect
atio
n
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas44
The geographic element is dominated by desert scruband desert forest and is characterised by a low temperaturein the coldest monthy = 382747 ndash17272 (CT) +00140 (DesertScrub) +00642 (DesertF)
All the segregations between chorotypes were strongThe strongest segregation occurs between chorotypes 1and 3 (DS = 06686 GS = 1406738 P lt 0001) followed bythe segregation between chorotypes 1 and 2 (DS = 06381GS = 586733 P lt 0001) and between chorotypes 2 and 3(DS = 03683 GS = 141756 P lt 0001)
The distributions of the three species that are not includ-ed in any chorotype are more similar to chorotype 1 than tothe other chorotypes (Figure 1) Two of them have a widedistribution throughout the continent the African GreyHornbill Tockus nasutus and the Red-billed Hornbill Tockuserythrorhynchus The other species is the Northern Ground-hornbill Bucorvus abyssinicus which occupies the sub-Saharan fringe from west to east
Discussion
The meaning of chorotypesThe majority of hornbill species in Africa (20 of the 23species) exhibit a well established discrete biogeographical
pattern constituted by three chorotypes with almost no over-lap occupying the western-central the eastern and thesouthern countries of Africa respectively Real et al (1997)considered that the fact that a large percentage of speciesconstitute chorotypes may be indicative of biogeographicalequilibrium in the biota analysed that is that species distri-butions have been well established for a long timeAccording to this interpretation hornbills would be a group ofspecies with long established distribution ranges in Africa
As a superspecies pair is shared between chorotypestwo and three in the form of Tockus flavirostris and T leu-
Figure 1 Dendrogram of classification of the distributions of Africanhornbills where the three significant chorotypes are signalled
Table 5 Segregations between groups of species on the dendro-gram nodes of Figure 1 The species of each chorotype and of theungrouped species appear in Appendix 1 DW(AxA) quantify theinternal homogeneity of each group GW(AxA) is a statisticalparameter obtained using a G test of independence ns = non sig-nificant Critical values for GW(AxA) are those of the Chi-squaredistribution with one degree of freedom
Group Cophenetic DW(AxA) GW(AxA) Significancecoefficient
Chorotype 1 0682 0557 47435 P lt 0001Chorotype 2 0870 0627 7227 P lt 001Chorotype 3 0560 0426 16429 P lt 0001Ungrouped species 0637 0415 1606 ns
Figure 2 Distribution of chorotype 1 with the number of species ofthe chorotype overlapping in each area
Figure 3 Distribution of chorotype 2 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 45
comelas these two chorotypes may be considered to bemore closely related to each other than to chorotype 1 asFigure 1 also suggests
The three chorotypes obtained are almost coincidentwith the Forest lsquoHornrsquo of Africa and Southern Savanna sub-regions respectively obtained by Williams et al (1999) for1911 species of Afrotropical birds (see also Crowe andCrowe 1982) This may indicate that this regionalisation notonly reflects the spatial responses of hornbills to environ-mental conditions but that these environmental factorsaffect the distributions of most Afrotropical birds at a conti-nental scale
The biotic element of the chorotypes should be distin-guished from ecological assemblages which imply the shar-ing of habitat or guilds which use the same trophicresources Chorotypes should rather be seen in a macroe-cological perspective (Brown 1995 Brown 1999) as emer-gent biogeographical structures generated by complex eco-logical systems which assemblages and guilds could beparticulate components of Chorotypes are defined here forlarge geographical units (countries) that include many differ-ent habitats and the biogeographic patterns defined on sobroad a scale are best characterised by macroenvironmen-tal factors that may go beyond the local habitat needs Abiotic element usually includes species that do not sharehabitats or way of life For example chorotype 2 includesTockus flavirostris and Tockus deckeni which are morecommon in open thorn and semi-arid savanna especially inCommiphora woodland but also includes Tockushemprichii that occupies rocky habitat in hills and gorgesDifferent ecological characteristics of habitat structure (suchas nest site availability or food abundance) may affect thehabitat selection of these species (see for example Hildeacuten1965 Elmberg et al 1994) but the macroenvironmental fac-
tors such as climate could create some basic conditionsthat are shared by the local environments so giving place toa common distribution pattern for species that are ecologi-cally different
Chorotypes also merit some consideration from the pointof view of conservation biology Birks (1987) proposed thatchorotypes could be valuable on the grounds of biogeo-graphical diversity So the degree to which each biotic ele-ment is represented in an area is a useful criterion for eval-uating its conservation interest combined with the richnessthe rarity and the conservation status of the species(Ramiacuterez and Vargas 1992 Turpie and Crowe 1994Williams et al 1996) and for assessing the representative-ness of the area (see Dasmann 1972 Austin and Margules1986 Margules 1986 Saeligtersdal and Birks 1993) Thiswould be useful for hornbill conservation planning on theAfrican scale although the selection of more specific sitesas particular reserves should require a more local and eco-logical approach
Particularly important in this respect are the areas withthe higher number of species of each chorotype which canbe seen in Figures 2 3 and 4 These may be considered asthe geographic cores of the chorotypes and may be spe-cially relevant areas for hornbills Even some disjoint distri-butions affect these areas For example Bycanistes breviswhich displays a patchy distribution due to the patchiness ofthe forest it occupies belongs to chorotype 3 but its north-ern distribution area patch is located in the core of chorotype2 (see Figures 3 and 4)
Macro-environmental and biotic characterisation of thechorotypesThe macro-environmental and biotic analyses of the hornbillchorotypes suggest a possible relationship of sexual dimor-phism with climatic seasonality and habitat type Hornbillsinhabiting non-seasonal areas supporting forest habitats(chorotype 1) have low sexual dimorphism while hornbillinhabiting seasonal savanna and grassland areas(chorotype 2) show greater levels of sexual dimorphismmales being bigger than females Bigger males might befavourably selected when competing for territories in thesavanna where suitable nesting sites are scarcer than in theforest
The species of chorotype 3 have a preference for wood-land or montane and coastal forest but not for rainforestand occur in the countries with colder winters and moredesert scrub and desert forest This could suggest that horn-bills can inhabit desert environments as long as there issome tree or scrub coverage although because of the limi-tation of the scale of this study more specific observationsof habitat usage are needed to assess this possibility
The ungrouped pattern of distributionThe three species that are not grouped in any chorotype arethose with the widest distribution in Africa specially theAfrican Grey Hornbill Tockus nasutus and the Red-billedHornbill Tockus erythrorhynchus several subspecies ofwhich can be found on both sides of the belt of tallBrachystegia or miombo woodland that stretches acrosscentral Africa (Kemp 2001) These two latter species are
Figure 4 Distribution of chorotype 3 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas46
also singled out because of their movements T nasutus hasbeen reported to make regular movements near the Saharaand irregular irruptive movements in southern Africa (Kemp2001) while T erythrorhynchus appears to undertake regu-lar movements all over its distribution area In contrast theNorthern Ground-hornbill Bucorvus abyssinicus resides inthe sub-Saharan African savannas north of the equator(Kemp 1995)
These three species are the only ones occurring inMauritania Mali Niger and Burkina Faso which representthe northernmost distribution for the hornbills in Africa andT nasutus is the only hornbill that is also present in theArabian Peninsula It is note-worthy that if only the Africandistribution of the African Grey Hornbill is considered thenthese three species are grouped together constituting afourth chorotype and then all African hornbill species followa discrete distribution pattern in a strictly African contextHowever the presence of Tockus nasutus in south-east Asiaseems to have somehow disrupted the well established dis-crete biogeographical configuration of the African hornbills
Acknowledgements mdash We are grateful to Dr Alan Kemp Dr PilaiPoonswad Dr Helen de Klerk and Dr Juan Carranza for their criti-cal comments on an early version of this paper This study was fund-ed by the European Commision and Spanish Government project1FD97ndash1571ndashC02ndash01
References
Austin MP and Margules CR 1986 Assessing representativenessIn Usher MB (ed) Wildlife Conservation Evaluation pp 45ndash68Chapman and Hall London
Austin MP and Smith TM 1989 A new model for the continuumconcept Vegetatio 83 35ndash47
Baroni-Urbani C and Buser MW 1976 Similarity of binary dataSystematic Zoology 25 251ndash259
Baroni-Urbani C Rufo S and Vigna-Taglianti A 1978 Materialiper una biogeografia italiana fondata su alguni generi dicoleotteri cicindelidi carabidi e crisomelidi Estratto dellaMemorie della Societa Entomologica Italiana 56 35ndash92
Birks HJB 1987 Recent methodological development in quantita-tive descriptive biogeography Annales Zoology Fennici 24165ndash178
Brown JH 1995 Macroecology The University of Chicago PressChicago
Brown JH 1999 Macroecology progress and prospect Oikos 873ndash14
Cordellier S and Didiot B 1999 Lrsquoeacutetet du monde Eacuteditions LaDeacutecouverte et Syros Pariacutes
Crowe TM and Crowe AA 1982 Patterns in distribution diversityand endemism in Afrotropical birds Journal of Zoology London198 417ndash442
Crowe TM and Kemp AC 1988 African historical biogeography asreflected by Galliform and Hornbill evolution In Ouellet H (ed)Acta XIX Congressus Internationalis Ornithologici Vol 2 pp2510ndash2518 University of Ottawa Press Ottawa
Dasmann RF 1972 Towards a system for classifying naturalregions of the world and their representation by National Parksand Reserves Biological Conservation 4 247ndash255
Elmberg J Sjoumlberg K Nummi P and Poumlysauml H 1994 Patterns oflake acidity and waterfowl communities Hydrobiologia 279280201ndash206
FAO 1984 Agroclimatological data for Africa FAO Plant Productionand Protection Series No 22 Vols 1 and 2 Rome
Guillet A and Crowe TM 1985 Patterns of distribution speciesrichness endemism and guild composition of water-birds inAfrica African Journal of Ecology 23 89ndash120
Guillet A and Crowe TM 1986 A preliminary investigation of pat-terns of distribution and species richness of southern Africanwaterbirds South African Journal of Wildlife Research 1665ndash81
Hengeveld R 1990 Dynamic Biogeography Cambridge UniversityPress Cambridge
Hengeveld R and Hogeweg P 1979 Cluster analysis of the distri-bution patterns of Dutch carabid species (Col) In Orloci L RaoCR and Stiteler WM (eds) Multivariate Methods in EcologicalWork pp 65ndash86 International Co-operative Publishing HouseBurtonsville
Hildeacuten O 1965 Habitat selection in birds a review AnnalesZoology Fennici 2 53ndash75
Hosmer DW and Lemeshow S 1989 Applied Logistic RegressionJohn Wiley and Sons Inc New York
Kemp A 1995 The Hornbills Bucerotiformes Oxford UniversityPress
Kemp A 2001 Family Bucerotidae (Hornbills) In Del Hoyo J ElliotA and Sargatal J (eds) Handbook of the Birds of the World Vol6 Mousebirds to Hornbills pp 436ndash523 Lynx Editions Barcelona
Kikkawa J and Pearse K 1969 Geographical distribution of landbirds in Australia a numerical analysis Australian Journal ofZoology 17 821ndash840
Lance GN and Williams WT 1967 A general theory of classificato-ry sorting strategies I Hierarchical systems Computer Journa l9 373ndash380
Margules CR 1986 Conservation evaluation in practice In UsherMB (ed) Wildlife Conservation Evaluation pp 298ndash314 Chapmanand Hall London
Maacuterquez AL Real R Vargas JM and Salvo AE 1997 On identify-ing common distribution patterns and their causal factors a prob-abilistic method applied to pteridophytes in the Iberian PeninsulaJournal of Biogeography 24 613ndash631
Mayr E 1965 What is a fauna Zoologisches Jahrbuch derSystematik 92 473ndash486
McCoy ED Bell SS and Walters K 1986 Identifying bioticboundaries along environmental gradients Ecology 67 749ndash759
Olivero J Real R and Vargas JM 1998 Distribution of breedingwintering and resident waterbirds in Europe biotic regions andthe macroclimate Ornis Fennica 75 153ndash175
Rabinovich JE and Rapoport EH 1975 Geographical variation ofdiversity in passerine birds Journal of Biogeography 2141ndash157
Ramiacuterez JM and Vargas JM 1992 Contribucioacuten de la biogeografiacuteaa la gestioacuten del medio ambiente y a la conservacioacuten de lasespecies In Vargas JM Real R and Antuacutenez A (eds) Objetivos yMeacutetodos Biogeograacuteficos Aplicaciones en Herpetologiacutea 2 pp95ndash106 Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Vargas JM and Guerrero JC 1992a Anaacutelisis biogeograacuteficode clasificacioacuten de aacutereas y especies In Vargas JM Real R andAntuacutenez A (ed) Objetivos y Meacutetodos Biogeograacuteficos Aplicacionesen Herpetologiacutea Monografias Herpetoloacutegica 2 pp 73ndash84Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Guerrero JC and Ramiacuterez JM 1992b Identificacioacuten defronteras bioacuteticas significativas para los anfibios en la cuencahidrograacutefica del Sur de Espantildea Dontildeana Acta Vertebrata 1953ndash70
Real R Pleguezuelos JM and Fahd S 1997 The distribution pat-terns of reptiles in the Riff region northern Morocco AfricanJournal of Ecology 35 312ndash325
Saeligtersdal M and Birks HJB 1993 Assessing the representative-ness of nature reserves using multivariate analysis vascularplants and breeding birds in deciduous forest western Norway
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 47
Biological Conservation 65 121ndash132Sibley CS and Monroe BE 1990 Distribution and Taxonomy of
Birds of the World Yale University Press New HavenSneath PHA and Sokal RR 1973 Numerical Taxonomy The
Principles and Practices of Numerical Classification FreemanSan Francisco
Sokal RR and Rohlf FJ 1981 Biometry The Principles andPractice of Statistics in Biological Research WH Freeman andCompany New York
Turner JRG Lennon JJ and Lawrenson JA 1988 British birdspecies distributions and the energy theory Nature 355 539ndash541
Turpie J K and Crowe TM 1994 Patterns of distribution diversityand endemism of larger African mammals South AfricanJournal of Zoology 29 19ndash32
Whitehead PJ Bowman DMJS and Tideman SC 1992Biogeographic patterns environmental correlates and conserva-tion of avifauna in the Northern Territory Australia Journal ofBiogeography 19 151ndash161
Wiens JA 1989 Spatial scaling in ecology Functional Ecology 3385ndash397
Williams PH De Klerk HM and Crowe TM 1999 Interpreting bio-geographical boundaries among Afrotropical birds spatial pat-terns in richness gradients and species replacement Journal ofBiogeography 26 459ndash474
Williams P Gibbons D Margules C Rebelo A Humphries C andPressey R 1996 A comparison of richness hotspots rarityhotspots and complementary areas for conserving diversity ofBritish birds Conservation Biology 10 155ndash174
World Resources Institute 1994 World Resources 1994ndash95 DataBase Diskette World Resources Institute Washington DC
Wright GH 1983 Species-energy theory an extension to species-area theory Oikos 41 496ndash506
Resumen
Se ha estudiado el aacuterea de distribucioacuten por paiacuteses de las 23 especies de caacutelaos africanos con el fin de identificar grupos dedistribuciones significativamente similares es decir corotipos y distribuciones que se reemplazan gradualmente Se hautilizado un meacutetodo probabiliacutestico de clasificacioacuten para distinguir entre ambos tipos de patrones biogeograacuteficos Una vezobtenidos los corotipos se analizoacute como se ven afectados por la influencia del clima el aacuterea la superficie cubierta por losdiferentes tipos de haacutebitat y las perturbaciones humanas en cada paiacutes Tambieacuten se comproboacute si las caracteriacutesticas propiasde cada especie entre las que se encuentran variables relacionadas con el haacutebitat la alimentacioacuten su biometriacutea y diferentesparaacutemetros metaboacutelicos influyen en los patrones biogeograacuteficos detectados Se han encontrado 3 corotipos que incluyen a20 especies El corotipo centro-occidental estaacute formado por 9 especies caracterizadas por la preferencia por haacutebitats debosque y por el bajo dimorfismo sexual y por 25 paiacuteses caracterizados por presentar un rango de temperaturas muy bajoEl corotipo oriental incluye a tres especies que se caracterizan por presentar haacutebitos alimenticios terrestres y un elevadodimorfismo sexual y a 7 paiacuteses caracterizados por mostrar gran parte de su superficie cubierta por sabanas y pastos Elcorotipo meridional agrupa la distribucioacuten de 8 especies que se caracterizan por su preferencia por bosques de montantildea ycosteros y zonas arboladas y a 18 paiacuteses con temperaturas bajas en el mes maacutes friacuteo y grandes superficies cubiertas pormatorrales y bosques de desierto Hay 3 especies que no se agrupan en corotipos y que siguen un reemplazamiento graduala lo largo del aacuterea de estudio
Chorotype 1African Pied Hornbill Tockus fasciatusPiping Hornbill Bycanistes fistulatorRed-billed Dwarf-hornbill Tockus camurusBlack-casqued Hornbill Ceratogymna atrataBrown-cheeked Hornbill Bycanistes cylindricusLong-tailed Hornbill Tropicranus albocristatusBlack Dwarf-hornbill Tockus hartlaubiGrey-cheeked Hornbill Bycanistes subcylindricusYellow-casqued Hornbill Ceratogymna elata
Chorotype 2Eastern Yellow-billed Hornbill Tockus flavirostrisVon der Deckenrsquos Hornbill Tockus deckeniHemprichrsquos Hornbill Tockus hemprichii
Appendix 1 Biotic elements of the three chorotypes obtained in Africa for the hornbills The corresponding geographic elements are shownin Figures 2 3 and 4 The species nomenclature follows Kemp (2001)
Chorotype 3Southern Ground-hornbill Bucorvus leadbeateriCrowned Hornbill Tockus alboterminatusPale-billed Hornbill Tockus pallidirostrisTrumpeter Hornbill Bycanistes bucinatorSilvery-cheeked Hornbill Bycanistes brevisBradfieldrsquos Hornbill Tockus bradfieldiSouthern Yellow-billed Hornbill Tockus leucomelasMonteirorsquos Hornbill Tockus monteiri
Ungrouped speciesAfrican Grey Hornbill Tockus nasutusRed Billed Hornbill Tockus erythrorhynchusNorthern Ground-hornbill Bucorvus abyssinicus
Received August 2001 accepted July 2002Editor WRJ Dean
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74(1amp2) 39ndash47Printed in South Africa mdash All rights reserved
Copyright copy NISC Pty Ltd
OSTRICHISSN 0030ndash6525
Biogeographical zonation of African hornbills and their biotic andgeographic characterisations
A Romaacuten Muntildeoz Raimundo Real Jesuacutes Olivero Ana L Maacuterquez Joseacute C Guerrero Silvia B Baacutercena
and J Mario VargasDepartamento de Biologiacutea Animal Facultad de Ciencias Universidad de Maacutelaga E-29071 Maacutelaga Spain
Corresponding author e-mail romanumaes
We studied the geographic ranges of the 23 African hornbill species using countries as distribution units to search for eithergroups of similar distributions (chorotypes) or gradual replacement of species as two different types of distribution patternsWe used a probabilistic classification method to distinguish between these two types of biogeographic patterns Then weanalysed the influence of climate area surface covered by different types of habitat and human disturbance in each coun-try as well as biotic features of the species involving the habitat preference feeding requirements and habits body size sex-ual dimorphism and metabolic parameters on the biogeographic patterns detected We found that 20 species were signifi-cantly classified into three chorotypes The western-central chorotype comprises nine species characterised by their prefer-ence for forest habitat and their low sexual body mass dimorphism These species occur in 25 countries characterised bytheir low range of temperatures The eastern chorotype includes three species characterised by their preference for feedingon the ground and greater dimorphism in mass between the sexes Seven countries characterised by their high values ofsavanna and grassland surface area constitute the geographic element of this chorotype The southern chorotype involvesthe distribution of eight species characterised by their preference for montane and coastal forests or woodlands Eighteencountries fall into this chorotype and are characterised by their low values of temperature in the coldest month and greatersurface area of desert scrub and desert forest The other three species replace each other mainly following a gradual patternwithin the study area (A Spanish translation of the abstract is provided on p 47)
Hornbills are a group of species that share many featuresfrom the taxonomical and the morphological points of viewThe most obvious characteristics are the long and decurvedbill and the projecting casque unique among avian familiesfeatures that assign them to the order Bucerotiformes(Sibley and Monroe 1990) They occupy an extensive geo-graphical range and from a biogeographic point of viewhave traditionally been considered as exhibiting merely twodisjoint distributions in sub-Saharan Africa and in south-eastAsia respectively (Kemp 1995 2001) The African speciesoccupy a wider range of habitats than the Asian speciesoccurring in a variety of environments from rainforest to aridsavanna The ecological complexity of the African hornbillspecies may result in complex biogeographical trends whicharise from the interactions of the species with their environ-ment and between themselves Because of this complexityit is difficult to integrate the present knowledge of the indi-vidual ecological properties of African hornbill species tounderstand their biogeographical trends It is preferable tofollow a synthetic approach that might account for someemergent properties of the distribution of hornbills and couldbe useful to understand the environmental and biotic influ-ences on the distribution of African hornbills as a whole
The geographical relationship among the distributionpatterns of several species may result in a continuum gradi-ent that is a gradual substitution of species in space or ina discrete pattern that is a set of similar distributions spa-
tially differentiated from other sets of similar distributions(see Mayr 1965 Austin and Smith 1989 Hengeveld 1990)These two kinds of patterns could coexist for differentgroups of species and it is necessary to test which type ofpattern is followed by each species
Baroni-Urbani et al (1978) called a group of similar dis-tributions that can be operatively distinguished from theother species ranges a lsquochorotypersquo Chorotypes have twodifferent components the geographic element which is thewhole area covered by any distribution of the chorotype andthe biotic element (sensu Birks 1987) which is the group ofspecies whose distributions belong to the same chorotypeIn this way chorotype and biotic element are related but aredifferent concepts that refer to discrete biogeographical pat-terns When chorotypes cannot be operatively obtained forsome species then they may be assumed to follow a con-tinuous pattern
Climate (specifically the availability of water and energy)is considered by many authors to be a major macro-envi-ronmental factor determining large scale biogeographicalpatterns (Hengeveld and Hogeweg 1979 Guillet and Crowe1986 Wiens 1989 Saeligtersdal and Birks 1993) For exam-ple the availability of energy has been found to explain abiogeographical regionalisation for waterbirds in Africa(Guillet and Crowe 1985) and the geographical trends in birdspecies richness in several areas (Rabinovich and Rapoport1975 Wright 1983 Turner et al 1988) whereas the avail-
Introduction
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas40
ability of water has also been found to explain the biogeo-graphical patterns of waterbirds in Africa (Guillet and Crowe1985 1986) and in Australia (Kikkawa and Pearse 1969Whitehead et al 1992) The types of habitat availablehuman influences the different biotic characteristics of thespecies and their ecological plasticity can also play a role inthe configuration of their actual distributions Only Croweand Kemp (1988) studied broad scale distribution patterns ofAfrican hornbill species although they did not search forchorotypes nor related the biogeographical patterns tomacroenvironmental characteristics
In this paper we analyse the chorotypes of African horn-bill species and characterise their distribution patternsaccording to the biotic features of the species and the cli-mate habitat characteristics and human influence on theirgeographic elements
Methods
The species and the study areaWe analysed the distribution of the hornbills in theAfrotropical region We considered the four African generawhich contain 23 species (Appendix 1) We used countriesas operative geographic units (OGUs) because hornbills areconspicuous birds well studied at this level of resolution sothat it is not likely that missing data affect the analysesInformation on the species occurrences in the OGUs weretaken from Kemp (1995) and Kemp (2001)
Probabilistic classification analysis to recognisechorotypesReal et al (1992a) applied the approach of McCoy et al(1986) to classification to develop a probabilistic procedurefor recognising chorotypes Following this approach weobtained a matrix of geographical similarities between thedistributions of each pair of species (a and b) using Baroni-Urbani and Buserrsquos (1976) index
where A is the number of countries where the species a ispresent B is the number of countries where the species b ispresent C is the number of countries where the two speciesa and b occur together and D is the number of countriesfrom which the two species a and b are absent This coeffi-cient takes into account shared absences that is countriesoutside the distribution area of both species where otherAfrican hornbills are present and so the similarities are con-sidered in the context of the whole study area and not onlyin the context of the subset of countries that constitute thetwo distributions compared (Real et al 1992b) Conjointabsences are not likely the result of uneven sampling whenusing countries as OGUs However this index gives moreimportance to shared presences and the possibility that twodistributions are considered similar only because of theirshared absences is avoided by multiplying shared absencesby shared occurrences (Baroni-Urbani and Buser 1976)
The grouping of hornbill distributions was made usingthe agglomerative method of classification UPGMA(Unweighted Pair-Group Method using Arithmetic Averages)
(Sneath and Sokal 1973) We prefered a group-averagemethod over a centroid technique because the latter do notyield monotonic results (Lance and Williams 1967 Sneathand Sokal 1973) The result of the classification was repre-sented as a dendrogram
We used the table of critical values in Baroni-Urbani andBuser (1976) to obtain values that are significantly more sim-ilar (+) or significantly more dissimilar (-) than would beexpected at random (Real et al 1992a Maacuterquez et al 1997)
We considered chorotypes to be those clusters that bestcombined the following characteristics a high proportion ofsignificant similarities (+) within the cluster a low proportionof significant dissimilarities (-) within the cluster and a lowproportion of significant similarities (+) between the distribu-tions of the cluster and the distributions of its most similarcluster The degree to which a distribution cluster A com-bines these conditions is provided by the parameterDW(AxA) (McCoy et al 1986) For the mathematical expan-sion of this parameter see Olivero et al (1998) In this waywe computed the DW(AxA) values for every branch of thedendrogram A cluster was considered a chorotype if a)DW(AxA) = 07071 that is the maximum value possible orb) DW(AxA) was positive higher than those of the otherclusters including the distributions involved and a G-test ofindependence (Sokal and Rohlf 1981) which yielded theparameter GW(AxA) showed that the proportion of the lsquo+rsquosigns within the cluster was significantly higher thanbetween it and its most similar cluster The distributions thatdid not fulfil these conditions were considered ungroupedand so they were considered to follow a continuous patternof gradual substitutions in space
The definition of a cluster as chorotype implies that thereis a significant segregation between it and its most similarcluster We also tested the type of segregation (strong orweak) between chorotypes by computing the parameter DS(see McCoy et al 1986) which increases when the propor-tion of significant dissimilarities (lsquo-rsquo signs) is higher betweenthe chorotypes and lower within each of them and the pro-portion of significant similarities is lower between thechorotypes The segregation between chorotypes wasstrong if a) DS = 07071 or b) DS was positive and a G testof independence which yielded the parameter GS showedthat the proportion of lsquo-rsquo signs was significantly higherbetween the chorotypes than within them else the segrega-tion was weak The parameters GS and GW(AxA) follow theChi-square distribution with one degree of freedom
Environmental and biotic characterisation of thechorotypesWe used 27 variables related to climate type of habitat andhuman influences to characterise the geographic element ofeach chorotype (Table 1) and 26 variables related to thehabitat preferences feeding guilds and habits body sizesexual dimorphism and metabolic parameters of the speciesto characterise the corresponding biotic element (Table 2)
We made these characterisations using a forward step-wise logistic regression (Hosmer and Lemeshow 1989) toobtain the variables that significantly increased the probabil-ity of any country or species to be part of the chorotypeaccording to the following equation
B = CxD ( ) + C
CxD ( ) + A + B - C
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 41
p = ey 1 + ey
where p is the probability e is the basis of the Napierian log-arithms and y is a linear combination of the variables (iethe logit equation) The statistical significance of the modelwas evaluated by means of a Chi-squared test and theparameters in the logistic function were estimated by maxi-mum likelihood The significance of variables within themodel was tested using the log likelihood ratio (LR) criterionie the significance of the change in log likelihood when thevariable is removed from the model
Results
The values of the Baroni-Urbani and Buserrsquos index betweenthe distributions of African hornbills are shown in Table 3The matrix of significant similarities between these distribu-tions is shown in Table 4
We identified three chorotypes for the African hornbills(Table 5 Figure 1) involving the distributions of twentyspecies whereas the distributions of the three species leftfollowed a continuous pattern of spatial substitution
Chorotype 1 The biotic element of this chorotype con-sists of nine species (Appendix 1) whose distributions occu-py central-western Africa (Figure 2) Both the preference forforest habitats and the smaller difference in sexual bodymass dimorphism characterise the species belonging to thechorotype according to the following logit equation
The geographic element on the other hand is charac-terised by a low annual temperature range according to thefollowing logit equation
y = 59347 - 08685 (TR)Chorotype 2 The biotic element is constituted by three
species (Appendix 1) all of them belonging to the genusTockus and distributed in eastern Africa (lsquohornrsquo of Africa)(Figure 3) The species of this chorotype have in common ahigh sexual body mass dimorphism
y = -1549 + 012 (Sexual body mass dimorph)Geographically chorotype 2 is dominated by large
expanses of savanna and grasslandy = -27816 + 00001 (SavGrass)Chorotype 3 The biotic element of this chorotype com-
prises eight species (Appendix 1) mainly distributed insouthern Africa (Figure 4) and is characterised by the habi-tat preference for montane and coastal forest and woodland
y = 86779 ndash 855 (Sexual body mass dimorph) + -9479 (Habitat savanna) +622 (Habitat montane and coastal forest) +2355 (Habitat woodland) +6501 (Habitat forest)
y = 395 +
-424 (Habitat savanna) +525 (Habitat montane and coastal forest) +525 (Habitat woodland) -626 (Habitat forest)
Table 1 Variables used to characterise the geographic element ofeach chorotype and their sources
Mean annual precipitation (mm) 1 PMean annual temperature (degC) 1 TColdest month mean temperature (degC) 1 CTWarmest month mean temperature (degC) 1 WTAnnual temperature range (degC) 1 TRTotal land area (ha) 2 AreaTropical land area (ha) 2 TropArid land area (ha) 2 AridSemi-arid land area (ha) 2 SAridHumid land area (ha) 2 HumidTotal forest area (ha) 2 ForestDry forest area (ha) 2 DryFMoist forest area (ha) 2 MoistFMangroves area (ha) 2 MangroveDesertScrub area (ha) 2 DesertScrubWetlandMarsh area (ha) 2 WetlandSavannahGrassland area (ha) 2 SavGrassDesert forest area (ha) 2 DesertFVery dry forest area (ha) 2 VeryDryFDry deciduous forest area (ha) 2 DryDFHill and montane forest area (ha) 2 HillMontFMoist deciduous forest area (ha) 2 MoistFRain forest area (ha) 2 RainFCropland area (ha) 2 CroplandAverage annual deforestation () 2 DeforestAnnual loss of SavannahGrassland () 2 SGLossPopulation density (Person km-2) 2 PopDensGrowth domestic product per capita 3 Gdp
Sources 1 FAO (1984) 2 World Resources Institute (1994) 3 Cordellierand Didiot (1999)
Table 2 Variables used to characterise the biotic element of eachchorotype
Habitat preference HFeeding preference FeedCarnivorous diet CFrugivorous diet FPreferential feeding location FLArboreal feeding location ArbTerrestrial feeding location TerBody mass of males BMassMBody mass of females BMassFBill length of males BillLCasque volume of males CasqueVPrimary area of males PAreaSecondary area of males SAreaWing size of males WingMWing size of females WingFTail size of males TailMTail size of females TailFTarsus length of males TarsusBasal metabolic rate BMRFlight metabolism FMBasal metabolic rateFlight metabolism FMBMR Linear wing loading LWLAspect ratio ARSexual body mass dimorphism SBMDSexual wing length dimorphism SWLDSexual bill length dimorphism SBLD
Source Kemp (1995)
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas42Ta
ble
3V
alu
es o
f th
e s
imila
rity
ind
ex o
f B
aro
ni-U
rba
ni a
nd B
use
r b
etw
ee
n th
e d
istr
ibut
ion
s of
th
e 2
3 A
fric
an
ho
rnb
ill s
peci
es
Sim
ilari
ty v
alu
es
bet
wee
n sp
ecie
s gr
oup
ed
into
a c
ho
roty
pe
are
en
clos
ed in
a b
ox
Spe
cies
num
ber
s a
s in
Ap
pend
ix 1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
231
00
00
60
60
667
069
00
69
00
64
50
605
06
250
594
05
940
65
40
568
016
20
33
90
15
40
204
02
700
000
00
000
00
00
446
044
60
397
211
00
00
854
043
00
43
00
30
50
356
03
310
305
03
050
50
80
349
051
80
59
60
30
80
404
04
040
273
03
730
15
20
259
030
80
273
221
000
050
40
50
40
36
80
377
03
960
368
03
680
47
10
439
047
00
55
60
33
20
441
04
920
347
04
590
20
40
332
038
60
347
11
000
10
00
08
71
08
750
899
08
710
871
07
25
075
30
172
02
50
02
50
02
140
138
01
720
146
02
03
025
00
250
017
22
10
00
08
71
08
750
899
08
710
871
07
25
075
30
172
02
50
02
50
02
140
138
01
720
146
02
03
025
00
250
017
27
10
00
09
460
973
09
440
944
07
85
063
60
226
03
18
03
20
02
760
182
02
270
192
02
69
032
00
320
022
73
10
000
974
09
460
946
07
91
065
10
270
03
47
02
90
02
500
163
02
030
172
02
40
029
00
290
020
34
10
000
973
09
730
76
10
674
021
40
30
40
30
50
263
01
720
215
01
820
25
40
305
030
50
215
51
000
09
440
73
00
698
022
60
31
80
32
00
276
01
820
227
01
920
26
90
320
032
00
227
61
000
07
30
069
80
226
02
57
03
20
02
760
000
02
270
192
02
69
021
50
215
022
78
10
00
059
20
454
05
13
03
74
03
970
326
02
030
172
02
40
037
40
451
039
19
100
00
000
00
00
00
00
00
000
000
00
000
000
00
00
000
00
000
000
013
100
00
85
30
73
00
861
07
190
678
08
210
46
30
254
037
30
393
141
00
00
64
60
774
07
740
515
06
660
39
30
571
064
60
595
151
00
00
869
06
810
588
07
160
53
20
000
039
30
422
161
000
08
430
610
07
950
42
20
323
046
30
490
171
000
04
900
709
00
00
057
80
681
061
018
10
000
856
07
78
000
00
000
000
019
10
000
62
80
000
000
00
000
201
00
00
000
000
00
000
121
000
090
00
839
101
000
094
811
100
0
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 43Ta
ble
4 M
atr
ix o
f si
gni
fican
t si
mila
ritie
s be
twe
en t
he
dis
trib
utio
ns
of
the
23
Afr
ica
n h
orn
bill
spe
cies
S
ign
ifica
nt
sim
ilari
ties
betw
een
sp
eci
es g
rou
ped
into
a c
hor
oty
pe a
re e
nclo
sed
in a
box
S
peci
es n
umb
ers
as
in A
ppe
ndix
1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
230
++
++
00
00
+0
--
--
--
--
00
021
+0
0-
0-
--
0-
00
-0
0-
0-
--
-22
00
00
00
00
00
0-
00
-0
--
0-
1+
++
++
++
+-
--
--
--
--
--
2+
++
++
++
--
--
--
--
--
-7
++
++
++
--
--
--
--
--
-3
++
++
+-
--
--
--
--
--
4+
++
+-
--
--
--
--
--
5+
++
--
--
--
--
--
-6
++
--
--
--
--
--
-8
00
00
0-
--
-0
00
9-
--
--
--
--
--
13+
++
++
+0
-0
014
++
+0
+0
0+
015
++
0+
0-
00
16+
0+
0-
00
170
+-
0+
018
++
--
-19
0-
--
20-
--
12+
+10
+11 +
S
imila
rity
sig
nifi
cant
ly h
igh
er t
han
ran
dom
exp
ecta
tion
-
Sim
ilarit
y si
gn
ifica
ntly
low
er t
han
ran
dom
exp
ecta
tion
0
S
imila
rity
with
out
sig
nifi
can
t d
iffer
ence
with
the
ra
ndom
exp
ect
atio
n
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas44
The geographic element is dominated by desert scruband desert forest and is characterised by a low temperaturein the coldest monthy = 382747 ndash17272 (CT) +00140 (DesertScrub) +00642 (DesertF)
All the segregations between chorotypes were strongThe strongest segregation occurs between chorotypes 1and 3 (DS = 06686 GS = 1406738 P lt 0001) followed bythe segregation between chorotypes 1 and 2 (DS = 06381GS = 586733 P lt 0001) and between chorotypes 2 and 3(DS = 03683 GS = 141756 P lt 0001)
The distributions of the three species that are not includ-ed in any chorotype are more similar to chorotype 1 than tothe other chorotypes (Figure 1) Two of them have a widedistribution throughout the continent the African GreyHornbill Tockus nasutus and the Red-billed Hornbill Tockuserythrorhynchus The other species is the Northern Ground-hornbill Bucorvus abyssinicus which occupies the sub-Saharan fringe from west to east
Discussion
The meaning of chorotypesThe majority of hornbill species in Africa (20 of the 23species) exhibit a well established discrete biogeographical
pattern constituted by three chorotypes with almost no over-lap occupying the western-central the eastern and thesouthern countries of Africa respectively Real et al (1997)considered that the fact that a large percentage of speciesconstitute chorotypes may be indicative of biogeographicalequilibrium in the biota analysed that is that species distri-butions have been well established for a long timeAccording to this interpretation hornbills would be a group ofspecies with long established distribution ranges in Africa
As a superspecies pair is shared between chorotypestwo and three in the form of Tockus flavirostris and T leu-
Figure 1 Dendrogram of classification of the distributions of Africanhornbills where the three significant chorotypes are signalled
Table 5 Segregations between groups of species on the dendro-gram nodes of Figure 1 The species of each chorotype and of theungrouped species appear in Appendix 1 DW(AxA) quantify theinternal homogeneity of each group GW(AxA) is a statisticalparameter obtained using a G test of independence ns = non sig-nificant Critical values for GW(AxA) are those of the Chi-squaredistribution with one degree of freedom
Group Cophenetic DW(AxA) GW(AxA) Significancecoefficient
Chorotype 1 0682 0557 47435 P lt 0001Chorotype 2 0870 0627 7227 P lt 001Chorotype 3 0560 0426 16429 P lt 0001Ungrouped species 0637 0415 1606 ns
Figure 2 Distribution of chorotype 1 with the number of species ofthe chorotype overlapping in each area
Figure 3 Distribution of chorotype 2 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 45
comelas these two chorotypes may be considered to bemore closely related to each other than to chorotype 1 asFigure 1 also suggests
The three chorotypes obtained are almost coincidentwith the Forest lsquoHornrsquo of Africa and Southern Savanna sub-regions respectively obtained by Williams et al (1999) for1911 species of Afrotropical birds (see also Crowe andCrowe 1982) This may indicate that this regionalisation notonly reflects the spatial responses of hornbills to environ-mental conditions but that these environmental factorsaffect the distributions of most Afrotropical birds at a conti-nental scale
The biotic element of the chorotypes should be distin-guished from ecological assemblages which imply the shar-ing of habitat or guilds which use the same trophicresources Chorotypes should rather be seen in a macroe-cological perspective (Brown 1995 Brown 1999) as emer-gent biogeographical structures generated by complex eco-logical systems which assemblages and guilds could beparticulate components of Chorotypes are defined here forlarge geographical units (countries) that include many differ-ent habitats and the biogeographic patterns defined on sobroad a scale are best characterised by macroenvironmen-tal factors that may go beyond the local habitat needs Abiotic element usually includes species that do not sharehabitats or way of life For example chorotype 2 includesTockus flavirostris and Tockus deckeni which are morecommon in open thorn and semi-arid savanna especially inCommiphora woodland but also includes Tockushemprichii that occupies rocky habitat in hills and gorgesDifferent ecological characteristics of habitat structure (suchas nest site availability or food abundance) may affect thehabitat selection of these species (see for example Hildeacuten1965 Elmberg et al 1994) but the macroenvironmental fac-
tors such as climate could create some basic conditionsthat are shared by the local environments so giving place toa common distribution pattern for species that are ecologi-cally different
Chorotypes also merit some consideration from the pointof view of conservation biology Birks (1987) proposed thatchorotypes could be valuable on the grounds of biogeo-graphical diversity So the degree to which each biotic ele-ment is represented in an area is a useful criterion for eval-uating its conservation interest combined with the richnessthe rarity and the conservation status of the species(Ramiacuterez and Vargas 1992 Turpie and Crowe 1994Williams et al 1996) and for assessing the representative-ness of the area (see Dasmann 1972 Austin and Margules1986 Margules 1986 Saeligtersdal and Birks 1993) Thiswould be useful for hornbill conservation planning on theAfrican scale although the selection of more specific sitesas particular reserves should require a more local and eco-logical approach
Particularly important in this respect are the areas withthe higher number of species of each chorotype which canbe seen in Figures 2 3 and 4 These may be considered asthe geographic cores of the chorotypes and may be spe-cially relevant areas for hornbills Even some disjoint distri-butions affect these areas For example Bycanistes breviswhich displays a patchy distribution due to the patchiness ofthe forest it occupies belongs to chorotype 3 but its north-ern distribution area patch is located in the core of chorotype2 (see Figures 3 and 4)
Macro-environmental and biotic characterisation of thechorotypesThe macro-environmental and biotic analyses of the hornbillchorotypes suggest a possible relationship of sexual dimor-phism with climatic seasonality and habitat type Hornbillsinhabiting non-seasonal areas supporting forest habitats(chorotype 1) have low sexual dimorphism while hornbillinhabiting seasonal savanna and grassland areas(chorotype 2) show greater levels of sexual dimorphismmales being bigger than females Bigger males might befavourably selected when competing for territories in thesavanna where suitable nesting sites are scarcer than in theforest
The species of chorotype 3 have a preference for wood-land or montane and coastal forest but not for rainforestand occur in the countries with colder winters and moredesert scrub and desert forest This could suggest that horn-bills can inhabit desert environments as long as there issome tree or scrub coverage although because of the limi-tation of the scale of this study more specific observationsof habitat usage are needed to assess this possibility
The ungrouped pattern of distributionThe three species that are not grouped in any chorotype arethose with the widest distribution in Africa specially theAfrican Grey Hornbill Tockus nasutus and the Red-billedHornbill Tockus erythrorhynchus several subspecies ofwhich can be found on both sides of the belt of tallBrachystegia or miombo woodland that stretches acrosscentral Africa (Kemp 2001) These two latter species are
Figure 4 Distribution of chorotype 3 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas46
also singled out because of their movements T nasutus hasbeen reported to make regular movements near the Saharaand irregular irruptive movements in southern Africa (Kemp2001) while T erythrorhynchus appears to undertake regu-lar movements all over its distribution area In contrast theNorthern Ground-hornbill Bucorvus abyssinicus resides inthe sub-Saharan African savannas north of the equator(Kemp 1995)
These three species are the only ones occurring inMauritania Mali Niger and Burkina Faso which representthe northernmost distribution for the hornbills in Africa andT nasutus is the only hornbill that is also present in theArabian Peninsula It is note-worthy that if only the Africandistribution of the African Grey Hornbill is considered thenthese three species are grouped together constituting afourth chorotype and then all African hornbill species followa discrete distribution pattern in a strictly African contextHowever the presence of Tockus nasutus in south-east Asiaseems to have somehow disrupted the well established dis-crete biogeographical configuration of the African hornbills
Acknowledgements mdash We are grateful to Dr Alan Kemp Dr PilaiPoonswad Dr Helen de Klerk and Dr Juan Carranza for their criti-cal comments on an early version of this paper This study was fund-ed by the European Commision and Spanish Government project1FD97ndash1571ndashC02ndash01
References
Austin MP and Margules CR 1986 Assessing representativenessIn Usher MB (ed) Wildlife Conservation Evaluation pp 45ndash68Chapman and Hall London
Austin MP and Smith TM 1989 A new model for the continuumconcept Vegetatio 83 35ndash47
Baroni-Urbani C and Buser MW 1976 Similarity of binary dataSystematic Zoology 25 251ndash259
Baroni-Urbani C Rufo S and Vigna-Taglianti A 1978 Materialiper una biogeografia italiana fondata su alguni generi dicoleotteri cicindelidi carabidi e crisomelidi Estratto dellaMemorie della Societa Entomologica Italiana 56 35ndash92
Birks HJB 1987 Recent methodological development in quantita-tive descriptive biogeography Annales Zoology Fennici 24165ndash178
Brown JH 1995 Macroecology The University of Chicago PressChicago
Brown JH 1999 Macroecology progress and prospect Oikos 873ndash14
Cordellier S and Didiot B 1999 Lrsquoeacutetet du monde Eacuteditions LaDeacutecouverte et Syros Pariacutes
Crowe TM and Crowe AA 1982 Patterns in distribution diversityand endemism in Afrotropical birds Journal of Zoology London198 417ndash442
Crowe TM and Kemp AC 1988 African historical biogeography asreflected by Galliform and Hornbill evolution In Ouellet H (ed)Acta XIX Congressus Internationalis Ornithologici Vol 2 pp2510ndash2518 University of Ottawa Press Ottawa
Dasmann RF 1972 Towards a system for classifying naturalregions of the world and their representation by National Parksand Reserves Biological Conservation 4 247ndash255
Elmberg J Sjoumlberg K Nummi P and Poumlysauml H 1994 Patterns oflake acidity and waterfowl communities Hydrobiologia 279280201ndash206
FAO 1984 Agroclimatological data for Africa FAO Plant Productionand Protection Series No 22 Vols 1 and 2 Rome
Guillet A and Crowe TM 1985 Patterns of distribution speciesrichness endemism and guild composition of water-birds inAfrica African Journal of Ecology 23 89ndash120
Guillet A and Crowe TM 1986 A preliminary investigation of pat-terns of distribution and species richness of southern Africanwaterbirds South African Journal of Wildlife Research 1665ndash81
Hengeveld R 1990 Dynamic Biogeography Cambridge UniversityPress Cambridge
Hengeveld R and Hogeweg P 1979 Cluster analysis of the distri-bution patterns of Dutch carabid species (Col) In Orloci L RaoCR and Stiteler WM (eds) Multivariate Methods in EcologicalWork pp 65ndash86 International Co-operative Publishing HouseBurtonsville
Hildeacuten O 1965 Habitat selection in birds a review AnnalesZoology Fennici 2 53ndash75
Hosmer DW and Lemeshow S 1989 Applied Logistic RegressionJohn Wiley and Sons Inc New York
Kemp A 1995 The Hornbills Bucerotiformes Oxford UniversityPress
Kemp A 2001 Family Bucerotidae (Hornbills) In Del Hoyo J ElliotA and Sargatal J (eds) Handbook of the Birds of the World Vol6 Mousebirds to Hornbills pp 436ndash523 Lynx Editions Barcelona
Kikkawa J and Pearse K 1969 Geographical distribution of landbirds in Australia a numerical analysis Australian Journal ofZoology 17 821ndash840
Lance GN and Williams WT 1967 A general theory of classificato-ry sorting strategies I Hierarchical systems Computer Journa l9 373ndash380
Margules CR 1986 Conservation evaluation in practice In UsherMB (ed) Wildlife Conservation Evaluation pp 298ndash314 Chapmanand Hall London
Maacuterquez AL Real R Vargas JM and Salvo AE 1997 On identify-ing common distribution patterns and their causal factors a prob-abilistic method applied to pteridophytes in the Iberian PeninsulaJournal of Biogeography 24 613ndash631
Mayr E 1965 What is a fauna Zoologisches Jahrbuch derSystematik 92 473ndash486
McCoy ED Bell SS and Walters K 1986 Identifying bioticboundaries along environmental gradients Ecology 67 749ndash759
Olivero J Real R and Vargas JM 1998 Distribution of breedingwintering and resident waterbirds in Europe biotic regions andthe macroclimate Ornis Fennica 75 153ndash175
Rabinovich JE and Rapoport EH 1975 Geographical variation ofdiversity in passerine birds Journal of Biogeography 2141ndash157
Ramiacuterez JM and Vargas JM 1992 Contribucioacuten de la biogeografiacuteaa la gestioacuten del medio ambiente y a la conservacioacuten de lasespecies In Vargas JM Real R and Antuacutenez A (eds) Objetivos yMeacutetodos Biogeograacuteficos Aplicaciones en Herpetologiacutea 2 pp95ndash106 Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Vargas JM and Guerrero JC 1992a Anaacutelisis biogeograacuteficode clasificacioacuten de aacutereas y especies In Vargas JM Real R andAntuacutenez A (ed) Objetivos y Meacutetodos Biogeograacuteficos Aplicacionesen Herpetologiacutea Monografias Herpetoloacutegica 2 pp 73ndash84Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Guerrero JC and Ramiacuterez JM 1992b Identificacioacuten defronteras bioacuteticas significativas para los anfibios en la cuencahidrograacutefica del Sur de Espantildea Dontildeana Acta Vertebrata 1953ndash70
Real R Pleguezuelos JM and Fahd S 1997 The distribution pat-terns of reptiles in the Riff region northern Morocco AfricanJournal of Ecology 35 312ndash325
Saeligtersdal M and Birks HJB 1993 Assessing the representative-ness of nature reserves using multivariate analysis vascularplants and breeding birds in deciduous forest western Norway
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 47
Biological Conservation 65 121ndash132Sibley CS and Monroe BE 1990 Distribution and Taxonomy of
Birds of the World Yale University Press New HavenSneath PHA and Sokal RR 1973 Numerical Taxonomy The
Principles and Practices of Numerical Classification FreemanSan Francisco
Sokal RR and Rohlf FJ 1981 Biometry The Principles andPractice of Statistics in Biological Research WH Freeman andCompany New York
Turner JRG Lennon JJ and Lawrenson JA 1988 British birdspecies distributions and the energy theory Nature 355 539ndash541
Turpie J K and Crowe TM 1994 Patterns of distribution diversityand endemism of larger African mammals South AfricanJournal of Zoology 29 19ndash32
Whitehead PJ Bowman DMJS and Tideman SC 1992Biogeographic patterns environmental correlates and conserva-tion of avifauna in the Northern Territory Australia Journal ofBiogeography 19 151ndash161
Wiens JA 1989 Spatial scaling in ecology Functional Ecology 3385ndash397
Williams PH De Klerk HM and Crowe TM 1999 Interpreting bio-geographical boundaries among Afrotropical birds spatial pat-terns in richness gradients and species replacement Journal ofBiogeography 26 459ndash474
Williams P Gibbons D Margules C Rebelo A Humphries C andPressey R 1996 A comparison of richness hotspots rarityhotspots and complementary areas for conserving diversity ofBritish birds Conservation Biology 10 155ndash174
World Resources Institute 1994 World Resources 1994ndash95 DataBase Diskette World Resources Institute Washington DC
Wright GH 1983 Species-energy theory an extension to species-area theory Oikos 41 496ndash506
Resumen
Se ha estudiado el aacuterea de distribucioacuten por paiacuteses de las 23 especies de caacutelaos africanos con el fin de identificar grupos dedistribuciones significativamente similares es decir corotipos y distribuciones que se reemplazan gradualmente Se hautilizado un meacutetodo probabiliacutestico de clasificacioacuten para distinguir entre ambos tipos de patrones biogeograacuteficos Una vezobtenidos los corotipos se analizoacute como se ven afectados por la influencia del clima el aacuterea la superficie cubierta por losdiferentes tipos de haacutebitat y las perturbaciones humanas en cada paiacutes Tambieacuten se comproboacute si las caracteriacutesticas propiasde cada especie entre las que se encuentran variables relacionadas con el haacutebitat la alimentacioacuten su biometriacutea y diferentesparaacutemetros metaboacutelicos influyen en los patrones biogeograacuteficos detectados Se han encontrado 3 corotipos que incluyen a20 especies El corotipo centro-occidental estaacute formado por 9 especies caracterizadas por la preferencia por haacutebitats debosque y por el bajo dimorfismo sexual y por 25 paiacuteses caracterizados por presentar un rango de temperaturas muy bajoEl corotipo oriental incluye a tres especies que se caracterizan por presentar haacutebitos alimenticios terrestres y un elevadodimorfismo sexual y a 7 paiacuteses caracterizados por mostrar gran parte de su superficie cubierta por sabanas y pastos Elcorotipo meridional agrupa la distribucioacuten de 8 especies que se caracterizan por su preferencia por bosques de montantildea ycosteros y zonas arboladas y a 18 paiacuteses con temperaturas bajas en el mes maacutes friacuteo y grandes superficies cubiertas pormatorrales y bosques de desierto Hay 3 especies que no se agrupan en corotipos y que siguen un reemplazamiento graduala lo largo del aacuterea de estudio
Chorotype 1African Pied Hornbill Tockus fasciatusPiping Hornbill Bycanistes fistulatorRed-billed Dwarf-hornbill Tockus camurusBlack-casqued Hornbill Ceratogymna atrataBrown-cheeked Hornbill Bycanistes cylindricusLong-tailed Hornbill Tropicranus albocristatusBlack Dwarf-hornbill Tockus hartlaubiGrey-cheeked Hornbill Bycanistes subcylindricusYellow-casqued Hornbill Ceratogymna elata
Chorotype 2Eastern Yellow-billed Hornbill Tockus flavirostrisVon der Deckenrsquos Hornbill Tockus deckeniHemprichrsquos Hornbill Tockus hemprichii
Appendix 1 Biotic elements of the three chorotypes obtained in Africa for the hornbills The corresponding geographic elements are shownin Figures 2 3 and 4 The species nomenclature follows Kemp (2001)
Chorotype 3Southern Ground-hornbill Bucorvus leadbeateriCrowned Hornbill Tockus alboterminatusPale-billed Hornbill Tockus pallidirostrisTrumpeter Hornbill Bycanistes bucinatorSilvery-cheeked Hornbill Bycanistes brevisBradfieldrsquos Hornbill Tockus bradfieldiSouthern Yellow-billed Hornbill Tockus leucomelasMonteirorsquos Hornbill Tockus monteiri
Ungrouped speciesAfrican Grey Hornbill Tockus nasutusRed Billed Hornbill Tockus erythrorhynchusNorthern Ground-hornbill Bucorvus abyssinicus
Received August 2001 accepted July 2002Editor WRJ Dean
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas40
ability of water has also been found to explain the biogeo-graphical patterns of waterbirds in Africa (Guillet and Crowe1985 1986) and in Australia (Kikkawa and Pearse 1969Whitehead et al 1992) The types of habitat availablehuman influences the different biotic characteristics of thespecies and their ecological plasticity can also play a role inthe configuration of their actual distributions Only Croweand Kemp (1988) studied broad scale distribution patterns ofAfrican hornbill species although they did not search forchorotypes nor related the biogeographical patterns tomacroenvironmental characteristics
In this paper we analyse the chorotypes of African horn-bill species and characterise their distribution patternsaccording to the biotic features of the species and the cli-mate habitat characteristics and human influence on theirgeographic elements
Methods
The species and the study areaWe analysed the distribution of the hornbills in theAfrotropical region We considered the four African generawhich contain 23 species (Appendix 1) We used countriesas operative geographic units (OGUs) because hornbills areconspicuous birds well studied at this level of resolution sothat it is not likely that missing data affect the analysesInformation on the species occurrences in the OGUs weretaken from Kemp (1995) and Kemp (2001)
Probabilistic classification analysis to recognisechorotypesReal et al (1992a) applied the approach of McCoy et al(1986) to classification to develop a probabilistic procedurefor recognising chorotypes Following this approach weobtained a matrix of geographical similarities between thedistributions of each pair of species (a and b) using Baroni-Urbani and Buserrsquos (1976) index
where A is the number of countries where the species a ispresent B is the number of countries where the species b ispresent C is the number of countries where the two speciesa and b occur together and D is the number of countriesfrom which the two species a and b are absent This coeffi-cient takes into account shared absences that is countriesoutside the distribution area of both species where otherAfrican hornbills are present and so the similarities are con-sidered in the context of the whole study area and not onlyin the context of the subset of countries that constitute thetwo distributions compared (Real et al 1992b) Conjointabsences are not likely the result of uneven sampling whenusing countries as OGUs However this index gives moreimportance to shared presences and the possibility that twodistributions are considered similar only because of theirshared absences is avoided by multiplying shared absencesby shared occurrences (Baroni-Urbani and Buser 1976)
The grouping of hornbill distributions was made usingthe agglomerative method of classification UPGMA(Unweighted Pair-Group Method using Arithmetic Averages)
(Sneath and Sokal 1973) We prefered a group-averagemethod over a centroid technique because the latter do notyield monotonic results (Lance and Williams 1967 Sneathand Sokal 1973) The result of the classification was repre-sented as a dendrogram
We used the table of critical values in Baroni-Urbani andBuser (1976) to obtain values that are significantly more sim-ilar (+) or significantly more dissimilar (-) than would beexpected at random (Real et al 1992a Maacuterquez et al 1997)
We considered chorotypes to be those clusters that bestcombined the following characteristics a high proportion ofsignificant similarities (+) within the cluster a low proportionof significant dissimilarities (-) within the cluster and a lowproportion of significant similarities (+) between the distribu-tions of the cluster and the distributions of its most similarcluster The degree to which a distribution cluster A com-bines these conditions is provided by the parameterDW(AxA) (McCoy et al 1986) For the mathematical expan-sion of this parameter see Olivero et al (1998) In this waywe computed the DW(AxA) values for every branch of thedendrogram A cluster was considered a chorotype if a)DW(AxA) = 07071 that is the maximum value possible orb) DW(AxA) was positive higher than those of the otherclusters including the distributions involved and a G-test ofindependence (Sokal and Rohlf 1981) which yielded theparameter GW(AxA) showed that the proportion of the lsquo+rsquosigns within the cluster was significantly higher thanbetween it and its most similar cluster The distributions thatdid not fulfil these conditions were considered ungroupedand so they were considered to follow a continuous patternof gradual substitutions in space
The definition of a cluster as chorotype implies that thereis a significant segregation between it and its most similarcluster We also tested the type of segregation (strong orweak) between chorotypes by computing the parameter DS(see McCoy et al 1986) which increases when the propor-tion of significant dissimilarities (lsquo-rsquo signs) is higher betweenthe chorotypes and lower within each of them and the pro-portion of significant similarities is lower between thechorotypes The segregation between chorotypes wasstrong if a) DS = 07071 or b) DS was positive and a G testof independence which yielded the parameter GS showedthat the proportion of lsquo-rsquo signs was significantly higherbetween the chorotypes than within them else the segrega-tion was weak The parameters GS and GW(AxA) follow theChi-square distribution with one degree of freedom
Environmental and biotic characterisation of thechorotypesWe used 27 variables related to climate type of habitat andhuman influences to characterise the geographic element ofeach chorotype (Table 1) and 26 variables related to thehabitat preferences feeding guilds and habits body sizesexual dimorphism and metabolic parameters of the speciesto characterise the corresponding biotic element (Table 2)
We made these characterisations using a forward step-wise logistic regression (Hosmer and Lemeshow 1989) toobtain the variables that significantly increased the probabil-ity of any country or species to be part of the chorotypeaccording to the following equation
B = CxD ( ) + C
CxD ( ) + A + B - C
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 41
p = ey 1 + ey
where p is the probability e is the basis of the Napierian log-arithms and y is a linear combination of the variables (iethe logit equation) The statistical significance of the modelwas evaluated by means of a Chi-squared test and theparameters in the logistic function were estimated by maxi-mum likelihood The significance of variables within themodel was tested using the log likelihood ratio (LR) criterionie the significance of the change in log likelihood when thevariable is removed from the model
Results
The values of the Baroni-Urbani and Buserrsquos index betweenthe distributions of African hornbills are shown in Table 3The matrix of significant similarities between these distribu-tions is shown in Table 4
We identified three chorotypes for the African hornbills(Table 5 Figure 1) involving the distributions of twentyspecies whereas the distributions of the three species leftfollowed a continuous pattern of spatial substitution
Chorotype 1 The biotic element of this chorotype con-sists of nine species (Appendix 1) whose distributions occu-py central-western Africa (Figure 2) Both the preference forforest habitats and the smaller difference in sexual bodymass dimorphism characterise the species belonging to thechorotype according to the following logit equation
The geographic element on the other hand is charac-terised by a low annual temperature range according to thefollowing logit equation
y = 59347 - 08685 (TR)Chorotype 2 The biotic element is constituted by three
species (Appendix 1) all of them belonging to the genusTockus and distributed in eastern Africa (lsquohornrsquo of Africa)(Figure 3) The species of this chorotype have in common ahigh sexual body mass dimorphism
y = -1549 + 012 (Sexual body mass dimorph)Geographically chorotype 2 is dominated by large
expanses of savanna and grasslandy = -27816 + 00001 (SavGrass)Chorotype 3 The biotic element of this chorotype com-
prises eight species (Appendix 1) mainly distributed insouthern Africa (Figure 4) and is characterised by the habi-tat preference for montane and coastal forest and woodland
y = 86779 ndash 855 (Sexual body mass dimorph) + -9479 (Habitat savanna) +622 (Habitat montane and coastal forest) +2355 (Habitat woodland) +6501 (Habitat forest)
y = 395 +
-424 (Habitat savanna) +525 (Habitat montane and coastal forest) +525 (Habitat woodland) -626 (Habitat forest)
Table 1 Variables used to characterise the geographic element ofeach chorotype and their sources
Mean annual precipitation (mm) 1 PMean annual temperature (degC) 1 TColdest month mean temperature (degC) 1 CTWarmest month mean temperature (degC) 1 WTAnnual temperature range (degC) 1 TRTotal land area (ha) 2 AreaTropical land area (ha) 2 TropArid land area (ha) 2 AridSemi-arid land area (ha) 2 SAridHumid land area (ha) 2 HumidTotal forest area (ha) 2 ForestDry forest area (ha) 2 DryFMoist forest area (ha) 2 MoistFMangroves area (ha) 2 MangroveDesertScrub area (ha) 2 DesertScrubWetlandMarsh area (ha) 2 WetlandSavannahGrassland area (ha) 2 SavGrassDesert forest area (ha) 2 DesertFVery dry forest area (ha) 2 VeryDryFDry deciduous forest area (ha) 2 DryDFHill and montane forest area (ha) 2 HillMontFMoist deciduous forest area (ha) 2 MoistFRain forest area (ha) 2 RainFCropland area (ha) 2 CroplandAverage annual deforestation () 2 DeforestAnnual loss of SavannahGrassland () 2 SGLossPopulation density (Person km-2) 2 PopDensGrowth domestic product per capita 3 Gdp
Sources 1 FAO (1984) 2 World Resources Institute (1994) 3 Cordellierand Didiot (1999)
Table 2 Variables used to characterise the biotic element of eachchorotype
Habitat preference HFeeding preference FeedCarnivorous diet CFrugivorous diet FPreferential feeding location FLArboreal feeding location ArbTerrestrial feeding location TerBody mass of males BMassMBody mass of females BMassFBill length of males BillLCasque volume of males CasqueVPrimary area of males PAreaSecondary area of males SAreaWing size of males WingMWing size of females WingFTail size of males TailMTail size of females TailFTarsus length of males TarsusBasal metabolic rate BMRFlight metabolism FMBasal metabolic rateFlight metabolism FMBMR Linear wing loading LWLAspect ratio ARSexual body mass dimorphism SBMDSexual wing length dimorphism SWLDSexual bill length dimorphism SBLD
Source Kemp (1995)
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas42Ta
ble
3V
alu
es o
f th
e s
imila
rity
ind
ex o
f B
aro
ni-U
rba
ni a
nd B
use
r b
etw
ee
n th
e d
istr
ibut
ion
s of
th
e 2
3 A
fric
an
ho
rnb
ill s
peci
es
Sim
ilari
ty v
alu
es
bet
wee
n sp
ecie
s gr
oup
ed
into
a c
ho
roty
pe
are
en
clos
ed in
a b
ox
Spe
cies
num
ber
s a
s in
Ap
pend
ix 1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
231
00
00
60
60
667
069
00
69
00
64
50
605
06
250
594
05
940
65
40
568
016
20
33
90
15
40
204
02
700
000
00
000
00
00
446
044
60
397
211
00
00
854
043
00
43
00
30
50
356
03
310
305
03
050
50
80
349
051
80
59
60
30
80
404
04
040
273
03
730
15
20
259
030
80
273
221
000
050
40
50
40
36
80
377
03
960
368
03
680
47
10
439
047
00
55
60
33
20
441
04
920
347
04
590
20
40
332
038
60
347
11
000
10
00
08
71
08
750
899
08
710
871
07
25
075
30
172
02
50
02
50
02
140
138
01
720
146
02
03
025
00
250
017
22
10
00
08
71
08
750
899
08
710
871
07
25
075
30
172
02
50
02
50
02
140
138
01
720
146
02
03
025
00
250
017
27
10
00
09
460
973
09
440
944
07
85
063
60
226
03
18
03
20
02
760
182
02
270
192
02
69
032
00
320
022
73
10
000
974
09
460
946
07
91
065
10
270
03
47
02
90
02
500
163
02
030
172
02
40
029
00
290
020
34
10
000
973
09
730
76
10
674
021
40
30
40
30
50
263
01
720
215
01
820
25
40
305
030
50
215
51
000
09
440
73
00
698
022
60
31
80
32
00
276
01
820
227
01
920
26
90
320
032
00
227
61
000
07
30
069
80
226
02
57
03
20
02
760
000
02
270
192
02
69
021
50
215
022
78
10
00
059
20
454
05
13
03
74
03
970
326
02
030
172
02
40
037
40
451
039
19
100
00
000
00
00
00
00
00
000
000
00
000
000
00
00
000
00
000
000
013
100
00
85
30
73
00
861
07
190
678
08
210
46
30
254
037
30
393
141
00
00
64
60
774
07
740
515
06
660
39
30
571
064
60
595
151
00
00
869
06
810
588
07
160
53
20
000
039
30
422
161
000
08
430
610
07
950
42
20
323
046
30
490
171
000
04
900
709
00
00
057
80
681
061
018
10
000
856
07
78
000
00
000
000
019
10
000
62
80
000
000
00
000
201
00
00
000
000
00
000
121
000
090
00
839
101
000
094
811
100
0
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 43Ta
ble
4 M
atr
ix o
f si
gni
fican
t si
mila
ritie
s be
twe
en t
he
dis
trib
utio
ns
of
the
23
Afr
ica
n h
orn
bill
spe
cies
S
ign
ifica
nt
sim
ilari
ties
betw
een
sp
eci
es g
rou
ped
into
a c
hor
oty
pe a
re e
nclo
sed
in a
box
S
peci
es n
umb
ers
as
in A
ppe
ndix
1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
230
++
++
00
00
+0
--
--
--
--
00
021
+0
0-
0-
--
0-
00
-0
0-
0-
--
-22
00
00
00
00
00
0-
00
-0
--
0-
1+
++
++
++
+-
--
--
--
--
--
2+
++
++
++
--
--
--
--
--
-7
++
++
++
--
--
--
--
--
-3
++
++
+-
--
--
--
--
--
4+
++
+-
--
--
--
--
--
5+
++
--
--
--
--
--
-6
++
--
--
--
--
--
-8
00
00
0-
--
-0
00
9-
--
--
--
--
--
13+
++
++
+0
-0
014
++
+0
+0
0+
015
++
0+
0-
00
16+
0+
0-
00
170
+-
0+
018
++
--
-19
0-
--
20-
--
12+
+10
+11 +
S
imila
rity
sig
nifi
cant
ly h
igh
er t
han
ran
dom
exp
ecta
tion
-
Sim
ilarit
y si
gn
ifica
ntly
low
er t
han
ran
dom
exp
ecta
tion
0
S
imila
rity
with
out
sig
nifi
can
t d
iffer
ence
with
the
ra
ndom
exp
ect
atio
n
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas44
The geographic element is dominated by desert scruband desert forest and is characterised by a low temperaturein the coldest monthy = 382747 ndash17272 (CT) +00140 (DesertScrub) +00642 (DesertF)
All the segregations between chorotypes were strongThe strongest segregation occurs between chorotypes 1and 3 (DS = 06686 GS = 1406738 P lt 0001) followed bythe segregation between chorotypes 1 and 2 (DS = 06381GS = 586733 P lt 0001) and between chorotypes 2 and 3(DS = 03683 GS = 141756 P lt 0001)
The distributions of the three species that are not includ-ed in any chorotype are more similar to chorotype 1 than tothe other chorotypes (Figure 1) Two of them have a widedistribution throughout the continent the African GreyHornbill Tockus nasutus and the Red-billed Hornbill Tockuserythrorhynchus The other species is the Northern Ground-hornbill Bucorvus abyssinicus which occupies the sub-Saharan fringe from west to east
Discussion
The meaning of chorotypesThe majority of hornbill species in Africa (20 of the 23species) exhibit a well established discrete biogeographical
pattern constituted by three chorotypes with almost no over-lap occupying the western-central the eastern and thesouthern countries of Africa respectively Real et al (1997)considered that the fact that a large percentage of speciesconstitute chorotypes may be indicative of biogeographicalequilibrium in the biota analysed that is that species distri-butions have been well established for a long timeAccording to this interpretation hornbills would be a group ofspecies with long established distribution ranges in Africa
As a superspecies pair is shared between chorotypestwo and three in the form of Tockus flavirostris and T leu-
Figure 1 Dendrogram of classification of the distributions of Africanhornbills where the three significant chorotypes are signalled
Table 5 Segregations between groups of species on the dendro-gram nodes of Figure 1 The species of each chorotype and of theungrouped species appear in Appendix 1 DW(AxA) quantify theinternal homogeneity of each group GW(AxA) is a statisticalparameter obtained using a G test of independence ns = non sig-nificant Critical values for GW(AxA) are those of the Chi-squaredistribution with one degree of freedom
Group Cophenetic DW(AxA) GW(AxA) Significancecoefficient
Chorotype 1 0682 0557 47435 P lt 0001Chorotype 2 0870 0627 7227 P lt 001Chorotype 3 0560 0426 16429 P lt 0001Ungrouped species 0637 0415 1606 ns
Figure 2 Distribution of chorotype 1 with the number of species ofthe chorotype overlapping in each area
Figure 3 Distribution of chorotype 2 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 45
comelas these two chorotypes may be considered to bemore closely related to each other than to chorotype 1 asFigure 1 also suggests
The three chorotypes obtained are almost coincidentwith the Forest lsquoHornrsquo of Africa and Southern Savanna sub-regions respectively obtained by Williams et al (1999) for1911 species of Afrotropical birds (see also Crowe andCrowe 1982) This may indicate that this regionalisation notonly reflects the spatial responses of hornbills to environ-mental conditions but that these environmental factorsaffect the distributions of most Afrotropical birds at a conti-nental scale
The biotic element of the chorotypes should be distin-guished from ecological assemblages which imply the shar-ing of habitat or guilds which use the same trophicresources Chorotypes should rather be seen in a macroe-cological perspective (Brown 1995 Brown 1999) as emer-gent biogeographical structures generated by complex eco-logical systems which assemblages and guilds could beparticulate components of Chorotypes are defined here forlarge geographical units (countries) that include many differ-ent habitats and the biogeographic patterns defined on sobroad a scale are best characterised by macroenvironmen-tal factors that may go beyond the local habitat needs Abiotic element usually includes species that do not sharehabitats or way of life For example chorotype 2 includesTockus flavirostris and Tockus deckeni which are morecommon in open thorn and semi-arid savanna especially inCommiphora woodland but also includes Tockushemprichii that occupies rocky habitat in hills and gorgesDifferent ecological characteristics of habitat structure (suchas nest site availability or food abundance) may affect thehabitat selection of these species (see for example Hildeacuten1965 Elmberg et al 1994) but the macroenvironmental fac-
tors such as climate could create some basic conditionsthat are shared by the local environments so giving place toa common distribution pattern for species that are ecologi-cally different
Chorotypes also merit some consideration from the pointof view of conservation biology Birks (1987) proposed thatchorotypes could be valuable on the grounds of biogeo-graphical diversity So the degree to which each biotic ele-ment is represented in an area is a useful criterion for eval-uating its conservation interest combined with the richnessthe rarity and the conservation status of the species(Ramiacuterez and Vargas 1992 Turpie and Crowe 1994Williams et al 1996) and for assessing the representative-ness of the area (see Dasmann 1972 Austin and Margules1986 Margules 1986 Saeligtersdal and Birks 1993) Thiswould be useful for hornbill conservation planning on theAfrican scale although the selection of more specific sitesas particular reserves should require a more local and eco-logical approach
Particularly important in this respect are the areas withthe higher number of species of each chorotype which canbe seen in Figures 2 3 and 4 These may be considered asthe geographic cores of the chorotypes and may be spe-cially relevant areas for hornbills Even some disjoint distri-butions affect these areas For example Bycanistes breviswhich displays a patchy distribution due to the patchiness ofthe forest it occupies belongs to chorotype 3 but its north-ern distribution area patch is located in the core of chorotype2 (see Figures 3 and 4)
Macro-environmental and biotic characterisation of thechorotypesThe macro-environmental and biotic analyses of the hornbillchorotypes suggest a possible relationship of sexual dimor-phism with climatic seasonality and habitat type Hornbillsinhabiting non-seasonal areas supporting forest habitats(chorotype 1) have low sexual dimorphism while hornbillinhabiting seasonal savanna and grassland areas(chorotype 2) show greater levels of sexual dimorphismmales being bigger than females Bigger males might befavourably selected when competing for territories in thesavanna where suitable nesting sites are scarcer than in theforest
The species of chorotype 3 have a preference for wood-land or montane and coastal forest but not for rainforestand occur in the countries with colder winters and moredesert scrub and desert forest This could suggest that horn-bills can inhabit desert environments as long as there issome tree or scrub coverage although because of the limi-tation of the scale of this study more specific observationsof habitat usage are needed to assess this possibility
The ungrouped pattern of distributionThe three species that are not grouped in any chorotype arethose with the widest distribution in Africa specially theAfrican Grey Hornbill Tockus nasutus and the Red-billedHornbill Tockus erythrorhynchus several subspecies ofwhich can be found on both sides of the belt of tallBrachystegia or miombo woodland that stretches acrosscentral Africa (Kemp 2001) These two latter species are
Figure 4 Distribution of chorotype 3 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas46
also singled out because of their movements T nasutus hasbeen reported to make regular movements near the Saharaand irregular irruptive movements in southern Africa (Kemp2001) while T erythrorhynchus appears to undertake regu-lar movements all over its distribution area In contrast theNorthern Ground-hornbill Bucorvus abyssinicus resides inthe sub-Saharan African savannas north of the equator(Kemp 1995)
These three species are the only ones occurring inMauritania Mali Niger and Burkina Faso which representthe northernmost distribution for the hornbills in Africa andT nasutus is the only hornbill that is also present in theArabian Peninsula It is note-worthy that if only the Africandistribution of the African Grey Hornbill is considered thenthese three species are grouped together constituting afourth chorotype and then all African hornbill species followa discrete distribution pattern in a strictly African contextHowever the presence of Tockus nasutus in south-east Asiaseems to have somehow disrupted the well established dis-crete biogeographical configuration of the African hornbills
Acknowledgements mdash We are grateful to Dr Alan Kemp Dr PilaiPoonswad Dr Helen de Klerk and Dr Juan Carranza for their criti-cal comments on an early version of this paper This study was fund-ed by the European Commision and Spanish Government project1FD97ndash1571ndashC02ndash01
References
Austin MP and Margules CR 1986 Assessing representativenessIn Usher MB (ed) Wildlife Conservation Evaluation pp 45ndash68Chapman and Hall London
Austin MP and Smith TM 1989 A new model for the continuumconcept Vegetatio 83 35ndash47
Baroni-Urbani C and Buser MW 1976 Similarity of binary dataSystematic Zoology 25 251ndash259
Baroni-Urbani C Rufo S and Vigna-Taglianti A 1978 Materialiper una biogeografia italiana fondata su alguni generi dicoleotteri cicindelidi carabidi e crisomelidi Estratto dellaMemorie della Societa Entomologica Italiana 56 35ndash92
Birks HJB 1987 Recent methodological development in quantita-tive descriptive biogeography Annales Zoology Fennici 24165ndash178
Brown JH 1995 Macroecology The University of Chicago PressChicago
Brown JH 1999 Macroecology progress and prospect Oikos 873ndash14
Cordellier S and Didiot B 1999 Lrsquoeacutetet du monde Eacuteditions LaDeacutecouverte et Syros Pariacutes
Crowe TM and Crowe AA 1982 Patterns in distribution diversityand endemism in Afrotropical birds Journal of Zoology London198 417ndash442
Crowe TM and Kemp AC 1988 African historical biogeography asreflected by Galliform and Hornbill evolution In Ouellet H (ed)Acta XIX Congressus Internationalis Ornithologici Vol 2 pp2510ndash2518 University of Ottawa Press Ottawa
Dasmann RF 1972 Towards a system for classifying naturalregions of the world and their representation by National Parksand Reserves Biological Conservation 4 247ndash255
Elmberg J Sjoumlberg K Nummi P and Poumlysauml H 1994 Patterns oflake acidity and waterfowl communities Hydrobiologia 279280201ndash206
FAO 1984 Agroclimatological data for Africa FAO Plant Productionand Protection Series No 22 Vols 1 and 2 Rome
Guillet A and Crowe TM 1985 Patterns of distribution speciesrichness endemism and guild composition of water-birds inAfrica African Journal of Ecology 23 89ndash120
Guillet A and Crowe TM 1986 A preliminary investigation of pat-terns of distribution and species richness of southern Africanwaterbirds South African Journal of Wildlife Research 1665ndash81
Hengeveld R 1990 Dynamic Biogeography Cambridge UniversityPress Cambridge
Hengeveld R and Hogeweg P 1979 Cluster analysis of the distri-bution patterns of Dutch carabid species (Col) In Orloci L RaoCR and Stiteler WM (eds) Multivariate Methods in EcologicalWork pp 65ndash86 International Co-operative Publishing HouseBurtonsville
Hildeacuten O 1965 Habitat selection in birds a review AnnalesZoology Fennici 2 53ndash75
Hosmer DW and Lemeshow S 1989 Applied Logistic RegressionJohn Wiley and Sons Inc New York
Kemp A 1995 The Hornbills Bucerotiformes Oxford UniversityPress
Kemp A 2001 Family Bucerotidae (Hornbills) In Del Hoyo J ElliotA and Sargatal J (eds) Handbook of the Birds of the World Vol6 Mousebirds to Hornbills pp 436ndash523 Lynx Editions Barcelona
Kikkawa J and Pearse K 1969 Geographical distribution of landbirds in Australia a numerical analysis Australian Journal ofZoology 17 821ndash840
Lance GN and Williams WT 1967 A general theory of classificato-ry sorting strategies I Hierarchical systems Computer Journa l9 373ndash380
Margules CR 1986 Conservation evaluation in practice In UsherMB (ed) Wildlife Conservation Evaluation pp 298ndash314 Chapmanand Hall London
Maacuterquez AL Real R Vargas JM and Salvo AE 1997 On identify-ing common distribution patterns and their causal factors a prob-abilistic method applied to pteridophytes in the Iberian PeninsulaJournal of Biogeography 24 613ndash631
Mayr E 1965 What is a fauna Zoologisches Jahrbuch derSystematik 92 473ndash486
McCoy ED Bell SS and Walters K 1986 Identifying bioticboundaries along environmental gradients Ecology 67 749ndash759
Olivero J Real R and Vargas JM 1998 Distribution of breedingwintering and resident waterbirds in Europe biotic regions andthe macroclimate Ornis Fennica 75 153ndash175
Rabinovich JE and Rapoport EH 1975 Geographical variation ofdiversity in passerine birds Journal of Biogeography 2141ndash157
Ramiacuterez JM and Vargas JM 1992 Contribucioacuten de la biogeografiacuteaa la gestioacuten del medio ambiente y a la conservacioacuten de lasespecies In Vargas JM Real R and Antuacutenez A (eds) Objetivos yMeacutetodos Biogeograacuteficos Aplicaciones en Herpetologiacutea 2 pp95ndash106 Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Vargas JM and Guerrero JC 1992a Anaacutelisis biogeograacuteficode clasificacioacuten de aacutereas y especies In Vargas JM Real R andAntuacutenez A (ed) Objetivos y Meacutetodos Biogeograacuteficos Aplicacionesen Herpetologiacutea Monografias Herpetoloacutegica 2 pp 73ndash84Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Guerrero JC and Ramiacuterez JM 1992b Identificacioacuten defronteras bioacuteticas significativas para los anfibios en la cuencahidrograacutefica del Sur de Espantildea Dontildeana Acta Vertebrata 1953ndash70
Real R Pleguezuelos JM and Fahd S 1997 The distribution pat-terns of reptiles in the Riff region northern Morocco AfricanJournal of Ecology 35 312ndash325
Saeligtersdal M and Birks HJB 1993 Assessing the representative-ness of nature reserves using multivariate analysis vascularplants and breeding birds in deciduous forest western Norway
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 47
Biological Conservation 65 121ndash132Sibley CS and Monroe BE 1990 Distribution and Taxonomy of
Birds of the World Yale University Press New HavenSneath PHA and Sokal RR 1973 Numerical Taxonomy The
Principles and Practices of Numerical Classification FreemanSan Francisco
Sokal RR and Rohlf FJ 1981 Biometry The Principles andPractice of Statistics in Biological Research WH Freeman andCompany New York
Turner JRG Lennon JJ and Lawrenson JA 1988 British birdspecies distributions and the energy theory Nature 355 539ndash541
Turpie J K and Crowe TM 1994 Patterns of distribution diversityand endemism of larger African mammals South AfricanJournal of Zoology 29 19ndash32
Whitehead PJ Bowman DMJS and Tideman SC 1992Biogeographic patterns environmental correlates and conserva-tion of avifauna in the Northern Territory Australia Journal ofBiogeography 19 151ndash161
Wiens JA 1989 Spatial scaling in ecology Functional Ecology 3385ndash397
Williams PH De Klerk HM and Crowe TM 1999 Interpreting bio-geographical boundaries among Afrotropical birds spatial pat-terns in richness gradients and species replacement Journal ofBiogeography 26 459ndash474
Williams P Gibbons D Margules C Rebelo A Humphries C andPressey R 1996 A comparison of richness hotspots rarityhotspots and complementary areas for conserving diversity ofBritish birds Conservation Biology 10 155ndash174
World Resources Institute 1994 World Resources 1994ndash95 DataBase Diskette World Resources Institute Washington DC
Wright GH 1983 Species-energy theory an extension to species-area theory Oikos 41 496ndash506
Resumen
Se ha estudiado el aacuterea de distribucioacuten por paiacuteses de las 23 especies de caacutelaos africanos con el fin de identificar grupos dedistribuciones significativamente similares es decir corotipos y distribuciones que se reemplazan gradualmente Se hautilizado un meacutetodo probabiliacutestico de clasificacioacuten para distinguir entre ambos tipos de patrones biogeograacuteficos Una vezobtenidos los corotipos se analizoacute como se ven afectados por la influencia del clima el aacuterea la superficie cubierta por losdiferentes tipos de haacutebitat y las perturbaciones humanas en cada paiacutes Tambieacuten se comproboacute si las caracteriacutesticas propiasde cada especie entre las que se encuentran variables relacionadas con el haacutebitat la alimentacioacuten su biometriacutea y diferentesparaacutemetros metaboacutelicos influyen en los patrones biogeograacuteficos detectados Se han encontrado 3 corotipos que incluyen a20 especies El corotipo centro-occidental estaacute formado por 9 especies caracterizadas por la preferencia por haacutebitats debosque y por el bajo dimorfismo sexual y por 25 paiacuteses caracterizados por presentar un rango de temperaturas muy bajoEl corotipo oriental incluye a tres especies que se caracterizan por presentar haacutebitos alimenticios terrestres y un elevadodimorfismo sexual y a 7 paiacuteses caracterizados por mostrar gran parte de su superficie cubierta por sabanas y pastos Elcorotipo meridional agrupa la distribucioacuten de 8 especies que se caracterizan por su preferencia por bosques de montantildea ycosteros y zonas arboladas y a 18 paiacuteses con temperaturas bajas en el mes maacutes friacuteo y grandes superficies cubiertas pormatorrales y bosques de desierto Hay 3 especies que no se agrupan en corotipos y que siguen un reemplazamiento graduala lo largo del aacuterea de estudio
Chorotype 1African Pied Hornbill Tockus fasciatusPiping Hornbill Bycanistes fistulatorRed-billed Dwarf-hornbill Tockus camurusBlack-casqued Hornbill Ceratogymna atrataBrown-cheeked Hornbill Bycanistes cylindricusLong-tailed Hornbill Tropicranus albocristatusBlack Dwarf-hornbill Tockus hartlaubiGrey-cheeked Hornbill Bycanistes subcylindricusYellow-casqued Hornbill Ceratogymna elata
Chorotype 2Eastern Yellow-billed Hornbill Tockus flavirostrisVon der Deckenrsquos Hornbill Tockus deckeniHemprichrsquos Hornbill Tockus hemprichii
Appendix 1 Biotic elements of the three chorotypes obtained in Africa for the hornbills The corresponding geographic elements are shownin Figures 2 3 and 4 The species nomenclature follows Kemp (2001)
Chorotype 3Southern Ground-hornbill Bucorvus leadbeateriCrowned Hornbill Tockus alboterminatusPale-billed Hornbill Tockus pallidirostrisTrumpeter Hornbill Bycanistes bucinatorSilvery-cheeked Hornbill Bycanistes brevisBradfieldrsquos Hornbill Tockus bradfieldiSouthern Yellow-billed Hornbill Tockus leucomelasMonteirorsquos Hornbill Tockus monteiri
Ungrouped speciesAfrican Grey Hornbill Tockus nasutusRed Billed Hornbill Tockus erythrorhynchusNorthern Ground-hornbill Bucorvus abyssinicus
Received August 2001 accepted July 2002Editor WRJ Dean
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 41
p = ey 1 + ey
where p is the probability e is the basis of the Napierian log-arithms and y is a linear combination of the variables (iethe logit equation) The statistical significance of the modelwas evaluated by means of a Chi-squared test and theparameters in the logistic function were estimated by maxi-mum likelihood The significance of variables within themodel was tested using the log likelihood ratio (LR) criterionie the significance of the change in log likelihood when thevariable is removed from the model
Results
The values of the Baroni-Urbani and Buserrsquos index betweenthe distributions of African hornbills are shown in Table 3The matrix of significant similarities between these distribu-tions is shown in Table 4
We identified three chorotypes for the African hornbills(Table 5 Figure 1) involving the distributions of twentyspecies whereas the distributions of the three species leftfollowed a continuous pattern of spatial substitution
Chorotype 1 The biotic element of this chorotype con-sists of nine species (Appendix 1) whose distributions occu-py central-western Africa (Figure 2) Both the preference forforest habitats and the smaller difference in sexual bodymass dimorphism characterise the species belonging to thechorotype according to the following logit equation
The geographic element on the other hand is charac-terised by a low annual temperature range according to thefollowing logit equation
y = 59347 - 08685 (TR)Chorotype 2 The biotic element is constituted by three
species (Appendix 1) all of them belonging to the genusTockus and distributed in eastern Africa (lsquohornrsquo of Africa)(Figure 3) The species of this chorotype have in common ahigh sexual body mass dimorphism
y = -1549 + 012 (Sexual body mass dimorph)Geographically chorotype 2 is dominated by large
expanses of savanna and grasslandy = -27816 + 00001 (SavGrass)Chorotype 3 The biotic element of this chorotype com-
prises eight species (Appendix 1) mainly distributed insouthern Africa (Figure 4) and is characterised by the habi-tat preference for montane and coastal forest and woodland
y = 86779 ndash 855 (Sexual body mass dimorph) + -9479 (Habitat savanna) +622 (Habitat montane and coastal forest) +2355 (Habitat woodland) +6501 (Habitat forest)
y = 395 +
-424 (Habitat savanna) +525 (Habitat montane and coastal forest) +525 (Habitat woodland) -626 (Habitat forest)
Table 1 Variables used to characterise the geographic element ofeach chorotype and their sources
Mean annual precipitation (mm) 1 PMean annual temperature (degC) 1 TColdest month mean temperature (degC) 1 CTWarmest month mean temperature (degC) 1 WTAnnual temperature range (degC) 1 TRTotal land area (ha) 2 AreaTropical land area (ha) 2 TropArid land area (ha) 2 AridSemi-arid land area (ha) 2 SAridHumid land area (ha) 2 HumidTotal forest area (ha) 2 ForestDry forest area (ha) 2 DryFMoist forest area (ha) 2 MoistFMangroves area (ha) 2 MangroveDesertScrub area (ha) 2 DesertScrubWetlandMarsh area (ha) 2 WetlandSavannahGrassland area (ha) 2 SavGrassDesert forest area (ha) 2 DesertFVery dry forest area (ha) 2 VeryDryFDry deciduous forest area (ha) 2 DryDFHill and montane forest area (ha) 2 HillMontFMoist deciduous forest area (ha) 2 MoistFRain forest area (ha) 2 RainFCropland area (ha) 2 CroplandAverage annual deforestation () 2 DeforestAnnual loss of SavannahGrassland () 2 SGLossPopulation density (Person km-2) 2 PopDensGrowth domestic product per capita 3 Gdp
Sources 1 FAO (1984) 2 World Resources Institute (1994) 3 Cordellierand Didiot (1999)
Table 2 Variables used to characterise the biotic element of eachchorotype
Habitat preference HFeeding preference FeedCarnivorous diet CFrugivorous diet FPreferential feeding location FLArboreal feeding location ArbTerrestrial feeding location TerBody mass of males BMassMBody mass of females BMassFBill length of males BillLCasque volume of males CasqueVPrimary area of males PAreaSecondary area of males SAreaWing size of males WingMWing size of females WingFTail size of males TailMTail size of females TailFTarsus length of males TarsusBasal metabolic rate BMRFlight metabolism FMBasal metabolic rateFlight metabolism FMBMR Linear wing loading LWLAspect ratio ARSexual body mass dimorphism SBMDSexual wing length dimorphism SWLDSexual bill length dimorphism SBLD
Source Kemp (1995)
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas42Ta
ble
3V
alu
es o
f th
e s
imila
rity
ind
ex o
f B
aro
ni-U
rba
ni a
nd B
use
r b
etw
ee
n th
e d
istr
ibut
ion
s of
th
e 2
3 A
fric
an
ho
rnb
ill s
peci
es
Sim
ilari
ty v
alu
es
bet
wee
n sp
ecie
s gr
oup
ed
into
a c
ho
roty
pe
are
en
clos
ed in
a b
ox
Spe
cies
num
ber
s a
s in
Ap
pend
ix 1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
231
00
00
60
60
667
069
00
69
00
64
50
605
06
250
594
05
940
65
40
568
016
20
33
90
15
40
204
02
700
000
00
000
00
00
446
044
60
397
211
00
00
854
043
00
43
00
30
50
356
03
310
305
03
050
50
80
349
051
80
59
60
30
80
404
04
040
273
03
730
15
20
259
030
80
273
221
000
050
40
50
40
36
80
377
03
960
368
03
680
47
10
439
047
00
55
60
33
20
441
04
920
347
04
590
20
40
332
038
60
347
11
000
10
00
08
71
08
750
899
08
710
871
07
25
075
30
172
02
50
02
50
02
140
138
01
720
146
02
03
025
00
250
017
22
10
00
08
71
08
750
899
08
710
871
07
25
075
30
172
02
50
02
50
02
140
138
01
720
146
02
03
025
00
250
017
27
10
00
09
460
973
09
440
944
07
85
063
60
226
03
18
03
20
02
760
182
02
270
192
02
69
032
00
320
022
73
10
000
974
09
460
946
07
91
065
10
270
03
47
02
90
02
500
163
02
030
172
02
40
029
00
290
020
34
10
000
973
09
730
76
10
674
021
40
30
40
30
50
263
01
720
215
01
820
25
40
305
030
50
215
51
000
09
440
73
00
698
022
60
31
80
32
00
276
01
820
227
01
920
26
90
320
032
00
227
61
000
07
30
069
80
226
02
57
03
20
02
760
000
02
270
192
02
69
021
50
215
022
78
10
00
059
20
454
05
13
03
74
03
970
326
02
030
172
02
40
037
40
451
039
19
100
00
000
00
00
00
00
00
000
000
00
000
000
00
00
000
00
000
000
013
100
00
85
30
73
00
861
07
190
678
08
210
46
30
254
037
30
393
141
00
00
64
60
774
07
740
515
06
660
39
30
571
064
60
595
151
00
00
869
06
810
588
07
160
53
20
000
039
30
422
161
000
08
430
610
07
950
42
20
323
046
30
490
171
000
04
900
709
00
00
057
80
681
061
018
10
000
856
07
78
000
00
000
000
019
10
000
62
80
000
000
00
000
201
00
00
000
000
00
000
121
000
090
00
839
101
000
094
811
100
0
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 43Ta
ble
4 M
atr
ix o
f si
gni
fican
t si
mila
ritie
s be
twe
en t
he
dis
trib
utio
ns
of
the
23
Afr
ica
n h
orn
bill
spe
cies
S
ign
ifica
nt
sim
ilari
ties
betw
een
sp
eci
es g
rou
ped
into
a c
hor
oty
pe a
re e
nclo
sed
in a
box
S
peci
es n
umb
ers
as
in A
ppe
ndix
1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
230
++
++
00
00
+0
--
--
--
--
00
021
+0
0-
0-
--
0-
00
-0
0-
0-
--
-22
00
00
00
00
00
0-
00
-0
--
0-
1+
++
++
++
+-
--
--
--
--
--
2+
++
++
++
--
--
--
--
--
-7
++
++
++
--
--
--
--
--
-3
++
++
+-
--
--
--
--
--
4+
++
+-
--
--
--
--
--
5+
++
--
--
--
--
--
-6
++
--
--
--
--
--
-8
00
00
0-
--
-0
00
9-
--
--
--
--
--
13+
++
++
+0
-0
014
++
+0
+0
0+
015
++
0+
0-
00
16+
0+
0-
00
170
+-
0+
018
++
--
-19
0-
--
20-
--
12+
+10
+11 +
S
imila
rity
sig
nifi
cant
ly h
igh
er t
han
ran
dom
exp
ecta
tion
-
Sim
ilarit
y si
gn
ifica
ntly
low
er t
han
ran
dom
exp
ecta
tion
0
S
imila
rity
with
out
sig
nifi
can
t d
iffer
ence
with
the
ra
ndom
exp
ect
atio
n
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas44
The geographic element is dominated by desert scruband desert forest and is characterised by a low temperaturein the coldest monthy = 382747 ndash17272 (CT) +00140 (DesertScrub) +00642 (DesertF)
All the segregations between chorotypes were strongThe strongest segregation occurs between chorotypes 1and 3 (DS = 06686 GS = 1406738 P lt 0001) followed bythe segregation between chorotypes 1 and 2 (DS = 06381GS = 586733 P lt 0001) and between chorotypes 2 and 3(DS = 03683 GS = 141756 P lt 0001)
The distributions of the three species that are not includ-ed in any chorotype are more similar to chorotype 1 than tothe other chorotypes (Figure 1) Two of them have a widedistribution throughout the continent the African GreyHornbill Tockus nasutus and the Red-billed Hornbill Tockuserythrorhynchus The other species is the Northern Ground-hornbill Bucorvus abyssinicus which occupies the sub-Saharan fringe from west to east
Discussion
The meaning of chorotypesThe majority of hornbill species in Africa (20 of the 23species) exhibit a well established discrete biogeographical
pattern constituted by three chorotypes with almost no over-lap occupying the western-central the eastern and thesouthern countries of Africa respectively Real et al (1997)considered that the fact that a large percentage of speciesconstitute chorotypes may be indicative of biogeographicalequilibrium in the biota analysed that is that species distri-butions have been well established for a long timeAccording to this interpretation hornbills would be a group ofspecies with long established distribution ranges in Africa
As a superspecies pair is shared between chorotypestwo and three in the form of Tockus flavirostris and T leu-
Figure 1 Dendrogram of classification of the distributions of Africanhornbills where the three significant chorotypes are signalled
Table 5 Segregations between groups of species on the dendro-gram nodes of Figure 1 The species of each chorotype and of theungrouped species appear in Appendix 1 DW(AxA) quantify theinternal homogeneity of each group GW(AxA) is a statisticalparameter obtained using a G test of independence ns = non sig-nificant Critical values for GW(AxA) are those of the Chi-squaredistribution with one degree of freedom
Group Cophenetic DW(AxA) GW(AxA) Significancecoefficient
Chorotype 1 0682 0557 47435 P lt 0001Chorotype 2 0870 0627 7227 P lt 001Chorotype 3 0560 0426 16429 P lt 0001Ungrouped species 0637 0415 1606 ns
Figure 2 Distribution of chorotype 1 with the number of species ofthe chorotype overlapping in each area
Figure 3 Distribution of chorotype 2 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 45
comelas these two chorotypes may be considered to bemore closely related to each other than to chorotype 1 asFigure 1 also suggests
The three chorotypes obtained are almost coincidentwith the Forest lsquoHornrsquo of Africa and Southern Savanna sub-regions respectively obtained by Williams et al (1999) for1911 species of Afrotropical birds (see also Crowe andCrowe 1982) This may indicate that this regionalisation notonly reflects the spatial responses of hornbills to environ-mental conditions but that these environmental factorsaffect the distributions of most Afrotropical birds at a conti-nental scale
The biotic element of the chorotypes should be distin-guished from ecological assemblages which imply the shar-ing of habitat or guilds which use the same trophicresources Chorotypes should rather be seen in a macroe-cological perspective (Brown 1995 Brown 1999) as emer-gent biogeographical structures generated by complex eco-logical systems which assemblages and guilds could beparticulate components of Chorotypes are defined here forlarge geographical units (countries) that include many differ-ent habitats and the biogeographic patterns defined on sobroad a scale are best characterised by macroenvironmen-tal factors that may go beyond the local habitat needs Abiotic element usually includes species that do not sharehabitats or way of life For example chorotype 2 includesTockus flavirostris and Tockus deckeni which are morecommon in open thorn and semi-arid savanna especially inCommiphora woodland but also includes Tockushemprichii that occupies rocky habitat in hills and gorgesDifferent ecological characteristics of habitat structure (suchas nest site availability or food abundance) may affect thehabitat selection of these species (see for example Hildeacuten1965 Elmberg et al 1994) but the macroenvironmental fac-
tors such as climate could create some basic conditionsthat are shared by the local environments so giving place toa common distribution pattern for species that are ecologi-cally different
Chorotypes also merit some consideration from the pointof view of conservation biology Birks (1987) proposed thatchorotypes could be valuable on the grounds of biogeo-graphical diversity So the degree to which each biotic ele-ment is represented in an area is a useful criterion for eval-uating its conservation interest combined with the richnessthe rarity and the conservation status of the species(Ramiacuterez and Vargas 1992 Turpie and Crowe 1994Williams et al 1996) and for assessing the representative-ness of the area (see Dasmann 1972 Austin and Margules1986 Margules 1986 Saeligtersdal and Birks 1993) Thiswould be useful for hornbill conservation planning on theAfrican scale although the selection of more specific sitesas particular reserves should require a more local and eco-logical approach
Particularly important in this respect are the areas withthe higher number of species of each chorotype which canbe seen in Figures 2 3 and 4 These may be considered asthe geographic cores of the chorotypes and may be spe-cially relevant areas for hornbills Even some disjoint distri-butions affect these areas For example Bycanistes breviswhich displays a patchy distribution due to the patchiness ofthe forest it occupies belongs to chorotype 3 but its north-ern distribution area patch is located in the core of chorotype2 (see Figures 3 and 4)
Macro-environmental and biotic characterisation of thechorotypesThe macro-environmental and biotic analyses of the hornbillchorotypes suggest a possible relationship of sexual dimor-phism with climatic seasonality and habitat type Hornbillsinhabiting non-seasonal areas supporting forest habitats(chorotype 1) have low sexual dimorphism while hornbillinhabiting seasonal savanna and grassland areas(chorotype 2) show greater levels of sexual dimorphismmales being bigger than females Bigger males might befavourably selected when competing for territories in thesavanna where suitable nesting sites are scarcer than in theforest
The species of chorotype 3 have a preference for wood-land or montane and coastal forest but not for rainforestand occur in the countries with colder winters and moredesert scrub and desert forest This could suggest that horn-bills can inhabit desert environments as long as there issome tree or scrub coverage although because of the limi-tation of the scale of this study more specific observationsof habitat usage are needed to assess this possibility
The ungrouped pattern of distributionThe three species that are not grouped in any chorotype arethose with the widest distribution in Africa specially theAfrican Grey Hornbill Tockus nasutus and the Red-billedHornbill Tockus erythrorhynchus several subspecies ofwhich can be found on both sides of the belt of tallBrachystegia or miombo woodland that stretches acrosscentral Africa (Kemp 2001) These two latter species are
Figure 4 Distribution of chorotype 3 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas46
also singled out because of their movements T nasutus hasbeen reported to make regular movements near the Saharaand irregular irruptive movements in southern Africa (Kemp2001) while T erythrorhynchus appears to undertake regu-lar movements all over its distribution area In contrast theNorthern Ground-hornbill Bucorvus abyssinicus resides inthe sub-Saharan African savannas north of the equator(Kemp 1995)
These three species are the only ones occurring inMauritania Mali Niger and Burkina Faso which representthe northernmost distribution for the hornbills in Africa andT nasutus is the only hornbill that is also present in theArabian Peninsula It is note-worthy that if only the Africandistribution of the African Grey Hornbill is considered thenthese three species are grouped together constituting afourth chorotype and then all African hornbill species followa discrete distribution pattern in a strictly African contextHowever the presence of Tockus nasutus in south-east Asiaseems to have somehow disrupted the well established dis-crete biogeographical configuration of the African hornbills
Acknowledgements mdash We are grateful to Dr Alan Kemp Dr PilaiPoonswad Dr Helen de Klerk and Dr Juan Carranza for their criti-cal comments on an early version of this paper This study was fund-ed by the European Commision and Spanish Government project1FD97ndash1571ndashC02ndash01
References
Austin MP and Margules CR 1986 Assessing representativenessIn Usher MB (ed) Wildlife Conservation Evaluation pp 45ndash68Chapman and Hall London
Austin MP and Smith TM 1989 A new model for the continuumconcept Vegetatio 83 35ndash47
Baroni-Urbani C and Buser MW 1976 Similarity of binary dataSystematic Zoology 25 251ndash259
Baroni-Urbani C Rufo S and Vigna-Taglianti A 1978 Materialiper una biogeografia italiana fondata su alguni generi dicoleotteri cicindelidi carabidi e crisomelidi Estratto dellaMemorie della Societa Entomologica Italiana 56 35ndash92
Birks HJB 1987 Recent methodological development in quantita-tive descriptive biogeography Annales Zoology Fennici 24165ndash178
Brown JH 1995 Macroecology The University of Chicago PressChicago
Brown JH 1999 Macroecology progress and prospect Oikos 873ndash14
Cordellier S and Didiot B 1999 Lrsquoeacutetet du monde Eacuteditions LaDeacutecouverte et Syros Pariacutes
Crowe TM and Crowe AA 1982 Patterns in distribution diversityand endemism in Afrotropical birds Journal of Zoology London198 417ndash442
Crowe TM and Kemp AC 1988 African historical biogeography asreflected by Galliform and Hornbill evolution In Ouellet H (ed)Acta XIX Congressus Internationalis Ornithologici Vol 2 pp2510ndash2518 University of Ottawa Press Ottawa
Dasmann RF 1972 Towards a system for classifying naturalregions of the world and their representation by National Parksand Reserves Biological Conservation 4 247ndash255
Elmberg J Sjoumlberg K Nummi P and Poumlysauml H 1994 Patterns oflake acidity and waterfowl communities Hydrobiologia 279280201ndash206
FAO 1984 Agroclimatological data for Africa FAO Plant Productionand Protection Series No 22 Vols 1 and 2 Rome
Guillet A and Crowe TM 1985 Patterns of distribution speciesrichness endemism and guild composition of water-birds inAfrica African Journal of Ecology 23 89ndash120
Guillet A and Crowe TM 1986 A preliminary investigation of pat-terns of distribution and species richness of southern Africanwaterbirds South African Journal of Wildlife Research 1665ndash81
Hengeveld R 1990 Dynamic Biogeography Cambridge UniversityPress Cambridge
Hengeveld R and Hogeweg P 1979 Cluster analysis of the distri-bution patterns of Dutch carabid species (Col) In Orloci L RaoCR and Stiteler WM (eds) Multivariate Methods in EcologicalWork pp 65ndash86 International Co-operative Publishing HouseBurtonsville
Hildeacuten O 1965 Habitat selection in birds a review AnnalesZoology Fennici 2 53ndash75
Hosmer DW and Lemeshow S 1989 Applied Logistic RegressionJohn Wiley and Sons Inc New York
Kemp A 1995 The Hornbills Bucerotiformes Oxford UniversityPress
Kemp A 2001 Family Bucerotidae (Hornbills) In Del Hoyo J ElliotA and Sargatal J (eds) Handbook of the Birds of the World Vol6 Mousebirds to Hornbills pp 436ndash523 Lynx Editions Barcelona
Kikkawa J and Pearse K 1969 Geographical distribution of landbirds in Australia a numerical analysis Australian Journal ofZoology 17 821ndash840
Lance GN and Williams WT 1967 A general theory of classificato-ry sorting strategies I Hierarchical systems Computer Journa l9 373ndash380
Margules CR 1986 Conservation evaluation in practice In UsherMB (ed) Wildlife Conservation Evaluation pp 298ndash314 Chapmanand Hall London
Maacuterquez AL Real R Vargas JM and Salvo AE 1997 On identify-ing common distribution patterns and their causal factors a prob-abilistic method applied to pteridophytes in the Iberian PeninsulaJournal of Biogeography 24 613ndash631
Mayr E 1965 What is a fauna Zoologisches Jahrbuch derSystematik 92 473ndash486
McCoy ED Bell SS and Walters K 1986 Identifying bioticboundaries along environmental gradients Ecology 67 749ndash759
Olivero J Real R and Vargas JM 1998 Distribution of breedingwintering and resident waterbirds in Europe biotic regions andthe macroclimate Ornis Fennica 75 153ndash175
Rabinovich JE and Rapoport EH 1975 Geographical variation ofdiversity in passerine birds Journal of Biogeography 2141ndash157
Ramiacuterez JM and Vargas JM 1992 Contribucioacuten de la biogeografiacuteaa la gestioacuten del medio ambiente y a la conservacioacuten de lasespecies In Vargas JM Real R and Antuacutenez A (eds) Objetivos yMeacutetodos Biogeograacuteficos Aplicaciones en Herpetologiacutea 2 pp95ndash106 Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Vargas JM and Guerrero JC 1992a Anaacutelisis biogeograacuteficode clasificacioacuten de aacutereas y especies In Vargas JM Real R andAntuacutenez A (ed) Objetivos y Meacutetodos Biogeograacuteficos Aplicacionesen Herpetologiacutea Monografias Herpetoloacutegica 2 pp 73ndash84Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Guerrero JC and Ramiacuterez JM 1992b Identificacioacuten defronteras bioacuteticas significativas para los anfibios en la cuencahidrograacutefica del Sur de Espantildea Dontildeana Acta Vertebrata 1953ndash70
Real R Pleguezuelos JM and Fahd S 1997 The distribution pat-terns of reptiles in the Riff region northern Morocco AfricanJournal of Ecology 35 312ndash325
Saeligtersdal M and Birks HJB 1993 Assessing the representative-ness of nature reserves using multivariate analysis vascularplants and breeding birds in deciduous forest western Norway
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 47
Biological Conservation 65 121ndash132Sibley CS and Monroe BE 1990 Distribution and Taxonomy of
Birds of the World Yale University Press New HavenSneath PHA and Sokal RR 1973 Numerical Taxonomy The
Principles and Practices of Numerical Classification FreemanSan Francisco
Sokal RR and Rohlf FJ 1981 Biometry The Principles andPractice of Statistics in Biological Research WH Freeman andCompany New York
Turner JRG Lennon JJ and Lawrenson JA 1988 British birdspecies distributions and the energy theory Nature 355 539ndash541
Turpie J K and Crowe TM 1994 Patterns of distribution diversityand endemism of larger African mammals South AfricanJournal of Zoology 29 19ndash32
Whitehead PJ Bowman DMJS and Tideman SC 1992Biogeographic patterns environmental correlates and conserva-tion of avifauna in the Northern Territory Australia Journal ofBiogeography 19 151ndash161
Wiens JA 1989 Spatial scaling in ecology Functional Ecology 3385ndash397
Williams PH De Klerk HM and Crowe TM 1999 Interpreting bio-geographical boundaries among Afrotropical birds spatial pat-terns in richness gradients and species replacement Journal ofBiogeography 26 459ndash474
Williams P Gibbons D Margules C Rebelo A Humphries C andPressey R 1996 A comparison of richness hotspots rarityhotspots and complementary areas for conserving diversity ofBritish birds Conservation Biology 10 155ndash174
World Resources Institute 1994 World Resources 1994ndash95 DataBase Diskette World Resources Institute Washington DC
Wright GH 1983 Species-energy theory an extension to species-area theory Oikos 41 496ndash506
Resumen
Se ha estudiado el aacuterea de distribucioacuten por paiacuteses de las 23 especies de caacutelaos africanos con el fin de identificar grupos dedistribuciones significativamente similares es decir corotipos y distribuciones que se reemplazan gradualmente Se hautilizado un meacutetodo probabiliacutestico de clasificacioacuten para distinguir entre ambos tipos de patrones biogeograacuteficos Una vezobtenidos los corotipos se analizoacute como se ven afectados por la influencia del clima el aacuterea la superficie cubierta por losdiferentes tipos de haacutebitat y las perturbaciones humanas en cada paiacutes Tambieacuten se comproboacute si las caracteriacutesticas propiasde cada especie entre las que se encuentran variables relacionadas con el haacutebitat la alimentacioacuten su biometriacutea y diferentesparaacutemetros metaboacutelicos influyen en los patrones biogeograacuteficos detectados Se han encontrado 3 corotipos que incluyen a20 especies El corotipo centro-occidental estaacute formado por 9 especies caracterizadas por la preferencia por haacutebitats debosque y por el bajo dimorfismo sexual y por 25 paiacuteses caracterizados por presentar un rango de temperaturas muy bajoEl corotipo oriental incluye a tres especies que se caracterizan por presentar haacutebitos alimenticios terrestres y un elevadodimorfismo sexual y a 7 paiacuteses caracterizados por mostrar gran parte de su superficie cubierta por sabanas y pastos Elcorotipo meridional agrupa la distribucioacuten de 8 especies que se caracterizan por su preferencia por bosques de montantildea ycosteros y zonas arboladas y a 18 paiacuteses con temperaturas bajas en el mes maacutes friacuteo y grandes superficies cubiertas pormatorrales y bosques de desierto Hay 3 especies que no se agrupan en corotipos y que siguen un reemplazamiento graduala lo largo del aacuterea de estudio
Chorotype 1African Pied Hornbill Tockus fasciatusPiping Hornbill Bycanistes fistulatorRed-billed Dwarf-hornbill Tockus camurusBlack-casqued Hornbill Ceratogymna atrataBrown-cheeked Hornbill Bycanistes cylindricusLong-tailed Hornbill Tropicranus albocristatusBlack Dwarf-hornbill Tockus hartlaubiGrey-cheeked Hornbill Bycanistes subcylindricusYellow-casqued Hornbill Ceratogymna elata
Chorotype 2Eastern Yellow-billed Hornbill Tockus flavirostrisVon der Deckenrsquos Hornbill Tockus deckeniHemprichrsquos Hornbill Tockus hemprichii
Appendix 1 Biotic elements of the three chorotypes obtained in Africa for the hornbills The corresponding geographic elements are shownin Figures 2 3 and 4 The species nomenclature follows Kemp (2001)
Chorotype 3Southern Ground-hornbill Bucorvus leadbeateriCrowned Hornbill Tockus alboterminatusPale-billed Hornbill Tockus pallidirostrisTrumpeter Hornbill Bycanistes bucinatorSilvery-cheeked Hornbill Bycanistes brevisBradfieldrsquos Hornbill Tockus bradfieldiSouthern Yellow-billed Hornbill Tockus leucomelasMonteirorsquos Hornbill Tockus monteiri
Ungrouped speciesAfrican Grey Hornbill Tockus nasutusRed Billed Hornbill Tockus erythrorhynchusNorthern Ground-hornbill Bucorvus abyssinicus
Received August 2001 accepted July 2002Editor WRJ Dean
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas42Ta
ble
3V
alu
es o
f th
e s
imila
rity
ind
ex o
f B
aro
ni-U
rba
ni a
nd B
use
r b
etw
ee
n th
e d
istr
ibut
ion
s of
th
e 2
3 A
fric
an
ho
rnb
ill s
peci
es
Sim
ilari
ty v
alu
es
bet
wee
n sp
ecie
s gr
oup
ed
into
a c
ho
roty
pe
are
en
clos
ed in
a b
ox
Spe
cies
num
ber
s a
s in
Ap
pend
ix 1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
231
00
00
60
60
667
069
00
69
00
64
50
605
06
250
594
05
940
65
40
568
016
20
33
90
15
40
204
02
700
000
00
000
00
00
446
044
60
397
211
00
00
854
043
00
43
00
30
50
356
03
310
305
03
050
50
80
349
051
80
59
60
30
80
404
04
040
273
03
730
15
20
259
030
80
273
221
000
050
40
50
40
36
80
377
03
960
368
03
680
47
10
439
047
00
55
60
33
20
441
04
920
347
04
590
20
40
332
038
60
347
11
000
10
00
08
71
08
750
899
08
710
871
07
25
075
30
172
02
50
02
50
02
140
138
01
720
146
02
03
025
00
250
017
22
10
00
08
71
08
750
899
08
710
871
07
25
075
30
172
02
50
02
50
02
140
138
01
720
146
02
03
025
00
250
017
27
10
00
09
460
973
09
440
944
07
85
063
60
226
03
18
03
20
02
760
182
02
270
192
02
69
032
00
320
022
73
10
000
974
09
460
946
07
91
065
10
270
03
47
02
90
02
500
163
02
030
172
02
40
029
00
290
020
34
10
000
973
09
730
76
10
674
021
40
30
40
30
50
263
01
720
215
01
820
25
40
305
030
50
215
51
000
09
440
73
00
698
022
60
31
80
32
00
276
01
820
227
01
920
26
90
320
032
00
227
61
000
07
30
069
80
226
02
57
03
20
02
760
000
02
270
192
02
69
021
50
215
022
78
10
00
059
20
454
05
13
03
74
03
970
326
02
030
172
02
40
037
40
451
039
19
100
00
000
00
00
00
00
00
000
000
00
000
000
00
00
000
00
000
000
013
100
00
85
30
73
00
861
07
190
678
08
210
46
30
254
037
30
393
141
00
00
64
60
774
07
740
515
06
660
39
30
571
064
60
595
151
00
00
869
06
810
588
07
160
53
20
000
039
30
422
161
000
08
430
610
07
950
42
20
323
046
30
490
171
000
04
900
709
00
00
057
80
681
061
018
10
000
856
07
78
000
00
000
000
019
10
000
62
80
000
000
00
000
201
00
00
000
000
00
000
121
000
090
00
839
101
000
094
811
100
0
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 43Ta
ble
4 M
atr
ix o
f si
gni
fican
t si
mila
ritie
s be
twe
en t
he
dis
trib
utio
ns
of
the
23
Afr
ica
n h
orn
bill
spe
cies
S
ign
ifica
nt
sim
ilari
ties
betw
een
sp
eci
es g
rou
ped
into
a c
hor
oty
pe a
re e
nclo
sed
in a
box
S
peci
es n
umb
ers
as
in A
ppe
ndix
1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
230
++
++
00
00
+0
--
--
--
--
00
021
+0
0-
0-
--
0-
00
-0
0-
0-
--
-22
00
00
00
00
00
0-
00
-0
--
0-
1+
++
++
++
+-
--
--
--
--
--
2+
++
++
++
--
--
--
--
--
-7
++
++
++
--
--
--
--
--
-3
++
++
+-
--
--
--
--
--
4+
++
+-
--
--
--
--
--
5+
++
--
--
--
--
--
-6
++
--
--
--
--
--
-8
00
00
0-
--
-0
00
9-
--
--
--
--
--
13+
++
++
+0
-0
014
++
+0
+0
0+
015
++
0+
0-
00
16+
0+
0-
00
170
+-
0+
018
++
--
-19
0-
--
20-
--
12+
+10
+11 +
S
imila
rity
sig
nifi
cant
ly h
igh
er t
han
ran
dom
exp
ecta
tion
-
Sim
ilarit
y si
gn
ifica
ntly
low
er t
han
ran
dom
exp
ecta
tion
0
S
imila
rity
with
out
sig
nifi
can
t d
iffer
ence
with
the
ra
ndom
exp
ect
atio
n
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas44
The geographic element is dominated by desert scruband desert forest and is characterised by a low temperaturein the coldest monthy = 382747 ndash17272 (CT) +00140 (DesertScrub) +00642 (DesertF)
All the segregations between chorotypes were strongThe strongest segregation occurs between chorotypes 1and 3 (DS = 06686 GS = 1406738 P lt 0001) followed bythe segregation between chorotypes 1 and 2 (DS = 06381GS = 586733 P lt 0001) and between chorotypes 2 and 3(DS = 03683 GS = 141756 P lt 0001)
The distributions of the three species that are not includ-ed in any chorotype are more similar to chorotype 1 than tothe other chorotypes (Figure 1) Two of them have a widedistribution throughout the continent the African GreyHornbill Tockus nasutus and the Red-billed Hornbill Tockuserythrorhynchus The other species is the Northern Ground-hornbill Bucorvus abyssinicus which occupies the sub-Saharan fringe from west to east
Discussion
The meaning of chorotypesThe majority of hornbill species in Africa (20 of the 23species) exhibit a well established discrete biogeographical
pattern constituted by three chorotypes with almost no over-lap occupying the western-central the eastern and thesouthern countries of Africa respectively Real et al (1997)considered that the fact that a large percentage of speciesconstitute chorotypes may be indicative of biogeographicalequilibrium in the biota analysed that is that species distri-butions have been well established for a long timeAccording to this interpretation hornbills would be a group ofspecies with long established distribution ranges in Africa
As a superspecies pair is shared between chorotypestwo and three in the form of Tockus flavirostris and T leu-
Figure 1 Dendrogram of classification of the distributions of Africanhornbills where the three significant chorotypes are signalled
Table 5 Segregations between groups of species on the dendro-gram nodes of Figure 1 The species of each chorotype and of theungrouped species appear in Appendix 1 DW(AxA) quantify theinternal homogeneity of each group GW(AxA) is a statisticalparameter obtained using a G test of independence ns = non sig-nificant Critical values for GW(AxA) are those of the Chi-squaredistribution with one degree of freedom
Group Cophenetic DW(AxA) GW(AxA) Significancecoefficient
Chorotype 1 0682 0557 47435 P lt 0001Chorotype 2 0870 0627 7227 P lt 001Chorotype 3 0560 0426 16429 P lt 0001Ungrouped species 0637 0415 1606 ns
Figure 2 Distribution of chorotype 1 with the number of species ofthe chorotype overlapping in each area
Figure 3 Distribution of chorotype 2 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 45
comelas these two chorotypes may be considered to bemore closely related to each other than to chorotype 1 asFigure 1 also suggests
The three chorotypes obtained are almost coincidentwith the Forest lsquoHornrsquo of Africa and Southern Savanna sub-regions respectively obtained by Williams et al (1999) for1911 species of Afrotropical birds (see also Crowe andCrowe 1982) This may indicate that this regionalisation notonly reflects the spatial responses of hornbills to environ-mental conditions but that these environmental factorsaffect the distributions of most Afrotropical birds at a conti-nental scale
The biotic element of the chorotypes should be distin-guished from ecological assemblages which imply the shar-ing of habitat or guilds which use the same trophicresources Chorotypes should rather be seen in a macroe-cological perspective (Brown 1995 Brown 1999) as emer-gent biogeographical structures generated by complex eco-logical systems which assemblages and guilds could beparticulate components of Chorotypes are defined here forlarge geographical units (countries) that include many differ-ent habitats and the biogeographic patterns defined on sobroad a scale are best characterised by macroenvironmen-tal factors that may go beyond the local habitat needs Abiotic element usually includes species that do not sharehabitats or way of life For example chorotype 2 includesTockus flavirostris and Tockus deckeni which are morecommon in open thorn and semi-arid savanna especially inCommiphora woodland but also includes Tockushemprichii that occupies rocky habitat in hills and gorgesDifferent ecological characteristics of habitat structure (suchas nest site availability or food abundance) may affect thehabitat selection of these species (see for example Hildeacuten1965 Elmberg et al 1994) but the macroenvironmental fac-
tors such as climate could create some basic conditionsthat are shared by the local environments so giving place toa common distribution pattern for species that are ecologi-cally different
Chorotypes also merit some consideration from the pointof view of conservation biology Birks (1987) proposed thatchorotypes could be valuable on the grounds of biogeo-graphical diversity So the degree to which each biotic ele-ment is represented in an area is a useful criterion for eval-uating its conservation interest combined with the richnessthe rarity and the conservation status of the species(Ramiacuterez and Vargas 1992 Turpie and Crowe 1994Williams et al 1996) and for assessing the representative-ness of the area (see Dasmann 1972 Austin and Margules1986 Margules 1986 Saeligtersdal and Birks 1993) Thiswould be useful for hornbill conservation planning on theAfrican scale although the selection of more specific sitesas particular reserves should require a more local and eco-logical approach
Particularly important in this respect are the areas withthe higher number of species of each chorotype which canbe seen in Figures 2 3 and 4 These may be considered asthe geographic cores of the chorotypes and may be spe-cially relevant areas for hornbills Even some disjoint distri-butions affect these areas For example Bycanistes breviswhich displays a patchy distribution due to the patchiness ofthe forest it occupies belongs to chorotype 3 but its north-ern distribution area patch is located in the core of chorotype2 (see Figures 3 and 4)
Macro-environmental and biotic characterisation of thechorotypesThe macro-environmental and biotic analyses of the hornbillchorotypes suggest a possible relationship of sexual dimor-phism with climatic seasonality and habitat type Hornbillsinhabiting non-seasonal areas supporting forest habitats(chorotype 1) have low sexual dimorphism while hornbillinhabiting seasonal savanna and grassland areas(chorotype 2) show greater levels of sexual dimorphismmales being bigger than females Bigger males might befavourably selected when competing for territories in thesavanna where suitable nesting sites are scarcer than in theforest
The species of chorotype 3 have a preference for wood-land or montane and coastal forest but not for rainforestand occur in the countries with colder winters and moredesert scrub and desert forest This could suggest that horn-bills can inhabit desert environments as long as there issome tree or scrub coverage although because of the limi-tation of the scale of this study more specific observationsof habitat usage are needed to assess this possibility
The ungrouped pattern of distributionThe three species that are not grouped in any chorotype arethose with the widest distribution in Africa specially theAfrican Grey Hornbill Tockus nasutus and the Red-billedHornbill Tockus erythrorhynchus several subspecies ofwhich can be found on both sides of the belt of tallBrachystegia or miombo woodland that stretches acrosscentral Africa (Kemp 2001) These two latter species are
Figure 4 Distribution of chorotype 3 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas46
also singled out because of their movements T nasutus hasbeen reported to make regular movements near the Saharaand irregular irruptive movements in southern Africa (Kemp2001) while T erythrorhynchus appears to undertake regu-lar movements all over its distribution area In contrast theNorthern Ground-hornbill Bucorvus abyssinicus resides inthe sub-Saharan African savannas north of the equator(Kemp 1995)
These three species are the only ones occurring inMauritania Mali Niger and Burkina Faso which representthe northernmost distribution for the hornbills in Africa andT nasutus is the only hornbill that is also present in theArabian Peninsula It is note-worthy that if only the Africandistribution of the African Grey Hornbill is considered thenthese three species are grouped together constituting afourth chorotype and then all African hornbill species followa discrete distribution pattern in a strictly African contextHowever the presence of Tockus nasutus in south-east Asiaseems to have somehow disrupted the well established dis-crete biogeographical configuration of the African hornbills
Acknowledgements mdash We are grateful to Dr Alan Kemp Dr PilaiPoonswad Dr Helen de Klerk and Dr Juan Carranza for their criti-cal comments on an early version of this paper This study was fund-ed by the European Commision and Spanish Government project1FD97ndash1571ndashC02ndash01
References
Austin MP and Margules CR 1986 Assessing representativenessIn Usher MB (ed) Wildlife Conservation Evaluation pp 45ndash68Chapman and Hall London
Austin MP and Smith TM 1989 A new model for the continuumconcept Vegetatio 83 35ndash47
Baroni-Urbani C and Buser MW 1976 Similarity of binary dataSystematic Zoology 25 251ndash259
Baroni-Urbani C Rufo S and Vigna-Taglianti A 1978 Materialiper una biogeografia italiana fondata su alguni generi dicoleotteri cicindelidi carabidi e crisomelidi Estratto dellaMemorie della Societa Entomologica Italiana 56 35ndash92
Birks HJB 1987 Recent methodological development in quantita-tive descriptive biogeography Annales Zoology Fennici 24165ndash178
Brown JH 1995 Macroecology The University of Chicago PressChicago
Brown JH 1999 Macroecology progress and prospect Oikos 873ndash14
Cordellier S and Didiot B 1999 Lrsquoeacutetet du monde Eacuteditions LaDeacutecouverte et Syros Pariacutes
Crowe TM and Crowe AA 1982 Patterns in distribution diversityand endemism in Afrotropical birds Journal of Zoology London198 417ndash442
Crowe TM and Kemp AC 1988 African historical biogeography asreflected by Galliform and Hornbill evolution In Ouellet H (ed)Acta XIX Congressus Internationalis Ornithologici Vol 2 pp2510ndash2518 University of Ottawa Press Ottawa
Dasmann RF 1972 Towards a system for classifying naturalregions of the world and their representation by National Parksand Reserves Biological Conservation 4 247ndash255
Elmberg J Sjoumlberg K Nummi P and Poumlysauml H 1994 Patterns oflake acidity and waterfowl communities Hydrobiologia 279280201ndash206
FAO 1984 Agroclimatological data for Africa FAO Plant Productionand Protection Series No 22 Vols 1 and 2 Rome
Guillet A and Crowe TM 1985 Patterns of distribution speciesrichness endemism and guild composition of water-birds inAfrica African Journal of Ecology 23 89ndash120
Guillet A and Crowe TM 1986 A preliminary investigation of pat-terns of distribution and species richness of southern Africanwaterbirds South African Journal of Wildlife Research 1665ndash81
Hengeveld R 1990 Dynamic Biogeography Cambridge UniversityPress Cambridge
Hengeveld R and Hogeweg P 1979 Cluster analysis of the distri-bution patterns of Dutch carabid species (Col) In Orloci L RaoCR and Stiteler WM (eds) Multivariate Methods in EcologicalWork pp 65ndash86 International Co-operative Publishing HouseBurtonsville
Hildeacuten O 1965 Habitat selection in birds a review AnnalesZoology Fennici 2 53ndash75
Hosmer DW and Lemeshow S 1989 Applied Logistic RegressionJohn Wiley and Sons Inc New York
Kemp A 1995 The Hornbills Bucerotiformes Oxford UniversityPress
Kemp A 2001 Family Bucerotidae (Hornbills) In Del Hoyo J ElliotA and Sargatal J (eds) Handbook of the Birds of the World Vol6 Mousebirds to Hornbills pp 436ndash523 Lynx Editions Barcelona
Kikkawa J and Pearse K 1969 Geographical distribution of landbirds in Australia a numerical analysis Australian Journal ofZoology 17 821ndash840
Lance GN and Williams WT 1967 A general theory of classificato-ry sorting strategies I Hierarchical systems Computer Journa l9 373ndash380
Margules CR 1986 Conservation evaluation in practice In UsherMB (ed) Wildlife Conservation Evaluation pp 298ndash314 Chapmanand Hall London
Maacuterquez AL Real R Vargas JM and Salvo AE 1997 On identify-ing common distribution patterns and their causal factors a prob-abilistic method applied to pteridophytes in the Iberian PeninsulaJournal of Biogeography 24 613ndash631
Mayr E 1965 What is a fauna Zoologisches Jahrbuch derSystematik 92 473ndash486
McCoy ED Bell SS and Walters K 1986 Identifying bioticboundaries along environmental gradients Ecology 67 749ndash759
Olivero J Real R and Vargas JM 1998 Distribution of breedingwintering and resident waterbirds in Europe biotic regions andthe macroclimate Ornis Fennica 75 153ndash175
Rabinovich JE and Rapoport EH 1975 Geographical variation ofdiversity in passerine birds Journal of Biogeography 2141ndash157
Ramiacuterez JM and Vargas JM 1992 Contribucioacuten de la biogeografiacuteaa la gestioacuten del medio ambiente y a la conservacioacuten de lasespecies In Vargas JM Real R and Antuacutenez A (eds) Objetivos yMeacutetodos Biogeograacuteficos Aplicaciones en Herpetologiacutea 2 pp95ndash106 Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Vargas JM and Guerrero JC 1992a Anaacutelisis biogeograacuteficode clasificacioacuten de aacutereas y especies In Vargas JM Real R andAntuacutenez A (ed) Objetivos y Meacutetodos Biogeograacuteficos Aplicacionesen Herpetologiacutea Monografias Herpetoloacutegica 2 pp 73ndash84Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Guerrero JC and Ramiacuterez JM 1992b Identificacioacuten defronteras bioacuteticas significativas para los anfibios en la cuencahidrograacutefica del Sur de Espantildea Dontildeana Acta Vertebrata 1953ndash70
Real R Pleguezuelos JM and Fahd S 1997 The distribution pat-terns of reptiles in the Riff region northern Morocco AfricanJournal of Ecology 35 312ndash325
Saeligtersdal M and Birks HJB 1993 Assessing the representative-ness of nature reserves using multivariate analysis vascularplants and breeding birds in deciduous forest western Norway
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 47
Biological Conservation 65 121ndash132Sibley CS and Monroe BE 1990 Distribution and Taxonomy of
Birds of the World Yale University Press New HavenSneath PHA and Sokal RR 1973 Numerical Taxonomy The
Principles and Practices of Numerical Classification FreemanSan Francisco
Sokal RR and Rohlf FJ 1981 Biometry The Principles andPractice of Statistics in Biological Research WH Freeman andCompany New York
Turner JRG Lennon JJ and Lawrenson JA 1988 British birdspecies distributions and the energy theory Nature 355 539ndash541
Turpie J K and Crowe TM 1994 Patterns of distribution diversityand endemism of larger African mammals South AfricanJournal of Zoology 29 19ndash32
Whitehead PJ Bowman DMJS and Tideman SC 1992Biogeographic patterns environmental correlates and conserva-tion of avifauna in the Northern Territory Australia Journal ofBiogeography 19 151ndash161
Wiens JA 1989 Spatial scaling in ecology Functional Ecology 3385ndash397
Williams PH De Klerk HM and Crowe TM 1999 Interpreting bio-geographical boundaries among Afrotropical birds spatial pat-terns in richness gradients and species replacement Journal ofBiogeography 26 459ndash474
Williams P Gibbons D Margules C Rebelo A Humphries C andPressey R 1996 A comparison of richness hotspots rarityhotspots and complementary areas for conserving diversity ofBritish birds Conservation Biology 10 155ndash174
World Resources Institute 1994 World Resources 1994ndash95 DataBase Diskette World Resources Institute Washington DC
Wright GH 1983 Species-energy theory an extension to species-area theory Oikos 41 496ndash506
Resumen
Se ha estudiado el aacuterea de distribucioacuten por paiacuteses de las 23 especies de caacutelaos africanos con el fin de identificar grupos dedistribuciones significativamente similares es decir corotipos y distribuciones que se reemplazan gradualmente Se hautilizado un meacutetodo probabiliacutestico de clasificacioacuten para distinguir entre ambos tipos de patrones biogeograacuteficos Una vezobtenidos los corotipos se analizoacute como se ven afectados por la influencia del clima el aacuterea la superficie cubierta por losdiferentes tipos de haacutebitat y las perturbaciones humanas en cada paiacutes Tambieacuten se comproboacute si las caracteriacutesticas propiasde cada especie entre las que se encuentran variables relacionadas con el haacutebitat la alimentacioacuten su biometriacutea y diferentesparaacutemetros metaboacutelicos influyen en los patrones biogeograacuteficos detectados Se han encontrado 3 corotipos que incluyen a20 especies El corotipo centro-occidental estaacute formado por 9 especies caracterizadas por la preferencia por haacutebitats debosque y por el bajo dimorfismo sexual y por 25 paiacuteses caracterizados por presentar un rango de temperaturas muy bajoEl corotipo oriental incluye a tres especies que se caracterizan por presentar haacutebitos alimenticios terrestres y un elevadodimorfismo sexual y a 7 paiacuteses caracterizados por mostrar gran parte de su superficie cubierta por sabanas y pastos Elcorotipo meridional agrupa la distribucioacuten de 8 especies que se caracterizan por su preferencia por bosques de montantildea ycosteros y zonas arboladas y a 18 paiacuteses con temperaturas bajas en el mes maacutes friacuteo y grandes superficies cubiertas pormatorrales y bosques de desierto Hay 3 especies que no se agrupan en corotipos y que siguen un reemplazamiento graduala lo largo del aacuterea de estudio
Chorotype 1African Pied Hornbill Tockus fasciatusPiping Hornbill Bycanistes fistulatorRed-billed Dwarf-hornbill Tockus camurusBlack-casqued Hornbill Ceratogymna atrataBrown-cheeked Hornbill Bycanistes cylindricusLong-tailed Hornbill Tropicranus albocristatusBlack Dwarf-hornbill Tockus hartlaubiGrey-cheeked Hornbill Bycanistes subcylindricusYellow-casqued Hornbill Ceratogymna elata
Chorotype 2Eastern Yellow-billed Hornbill Tockus flavirostrisVon der Deckenrsquos Hornbill Tockus deckeniHemprichrsquos Hornbill Tockus hemprichii
Appendix 1 Biotic elements of the three chorotypes obtained in Africa for the hornbills The corresponding geographic elements are shownin Figures 2 3 and 4 The species nomenclature follows Kemp (2001)
Chorotype 3Southern Ground-hornbill Bucorvus leadbeateriCrowned Hornbill Tockus alboterminatusPale-billed Hornbill Tockus pallidirostrisTrumpeter Hornbill Bycanistes bucinatorSilvery-cheeked Hornbill Bycanistes brevisBradfieldrsquos Hornbill Tockus bradfieldiSouthern Yellow-billed Hornbill Tockus leucomelasMonteirorsquos Hornbill Tockus monteiri
Ungrouped speciesAfrican Grey Hornbill Tockus nasutusRed Billed Hornbill Tockus erythrorhynchusNorthern Ground-hornbill Bucorvus abyssinicus
Received August 2001 accepted July 2002Editor WRJ Dean
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 43Ta
ble
4 M
atr
ix o
f si
gni
fican
t si
mila
ritie
s be
twe
en t
he
dis
trib
utio
ns
of
the
23
Afr
ica
n h
orn
bill
spe
cies
S
ign
ifica
nt
sim
ilari
ties
betw
een
sp
eci
es g
rou
ped
into
a c
hor
oty
pe a
re e
nclo
sed
in a
box
S
peci
es n
umb
ers
as
in A
ppe
ndix
1
Spe
cies
2321
221
27
34
56
89
1314
1516
1718
1920
1210
11
230
++
++
00
00
+0
--
--
--
--
00
021
+0
0-
0-
--
0-
00
-0
0-
0-
--
-22
00
00
00
00
00
0-
00
-0
--
0-
1+
++
++
++
+-
--
--
--
--
--
2+
++
++
++
--
--
--
--
--
-7
++
++
++
--
--
--
--
--
-3
++
++
+-
--
--
--
--
--
4+
++
+-
--
--
--
--
--
5+
++
--
--
--
--
--
-6
++
--
--
--
--
--
-8
00
00
0-
--
-0
00
9-
--
--
--
--
--
13+
++
++
+0
-0
014
++
+0
+0
0+
015
++
0+
0-
00
16+
0+
0-
00
170
+-
0+
018
++
--
-19
0-
--
20-
--
12+
+10
+11 +
S
imila
rity
sig
nifi
cant
ly h
igh
er t
han
ran
dom
exp
ecta
tion
-
Sim
ilarit
y si
gn
ifica
ntly
low
er t
han
ran
dom
exp
ecta
tion
0
S
imila
rity
with
out
sig
nifi
can
t d
iffer
ence
with
the
ra
ndom
exp
ect
atio
n
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas44
The geographic element is dominated by desert scruband desert forest and is characterised by a low temperaturein the coldest monthy = 382747 ndash17272 (CT) +00140 (DesertScrub) +00642 (DesertF)
All the segregations between chorotypes were strongThe strongest segregation occurs between chorotypes 1and 3 (DS = 06686 GS = 1406738 P lt 0001) followed bythe segregation between chorotypes 1 and 2 (DS = 06381GS = 586733 P lt 0001) and between chorotypes 2 and 3(DS = 03683 GS = 141756 P lt 0001)
The distributions of the three species that are not includ-ed in any chorotype are more similar to chorotype 1 than tothe other chorotypes (Figure 1) Two of them have a widedistribution throughout the continent the African GreyHornbill Tockus nasutus and the Red-billed Hornbill Tockuserythrorhynchus The other species is the Northern Ground-hornbill Bucorvus abyssinicus which occupies the sub-Saharan fringe from west to east
Discussion
The meaning of chorotypesThe majority of hornbill species in Africa (20 of the 23species) exhibit a well established discrete biogeographical
pattern constituted by three chorotypes with almost no over-lap occupying the western-central the eastern and thesouthern countries of Africa respectively Real et al (1997)considered that the fact that a large percentage of speciesconstitute chorotypes may be indicative of biogeographicalequilibrium in the biota analysed that is that species distri-butions have been well established for a long timeAccording to this interpretation hornbills would be a group ofspecies with long established distribution ranges in Africa
As a superspecies pair is shared between chorotypestwo and three in the form of Tockus flavirostris and T leu-
Figure 1 Dendrogram of classification of the distributions of Africanhornbills where the three significant chorotypes are signalled
Table 5 Segregations between groups of species on the dendro-gram nodes of Figure 1 The species of each chorotype and of theungrouped species appear in Appendix 1 DW(AxA) quantify theinternal homogeneity of each group GW(AxA) is a statisticalparameter obtained using a G test of independence ns = non sig-nificant Critical values for GW(AxA) are those of the Chi-squaredistribution with one degree of freedom
Group Cophenetic DW(AxA) GW(AxA) Significancecoefficient
Chorotype 1 0682 0557 47435 P lt 0001Chorotype 2 0870 0627 7227 P lt 001Chorotype 3 0560 0426 16429 P lt 0001Ungrouped species 0637 0415 1606 ns
Figure 2 Distribution of chorotype 1 with the number of species ofthe chorotype overlapping in each area
Figure 3 Distribution of chorotype 2 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 45
comelas these two chorotypes may be considered to bemore closely related to each other than to chorotype 1 asFigure 1 also suggests
The three chorotypes obtained are almost coincidentwith the Forest lsquoHornrsquo of Africa and Southern Savanna sub-regions respectively obtained by Williams et al (1999) for1911 species of Afrotropical birds (see also Crowe andCrowe 1982) This may indicate that this regionalisation notonly reflects the spatial responses of hornbills to environ-mental conditions but that these environmental factorsaffect the distributions of most Afrotropical birds at a conti-nental scale
The biotic element of the chorotypes should be distin-guished from ecological assemblages which imply the shar-ing of habitat or guilds which use the same trophicresources Chorotypes should rather be seen in a macroe-cological perspective (Brown 1995 Brown 1999) as emer-gent biogeographical structures generated by complex eco-logical systems which assemblages and guilds could beparticulate components of Chorotypes are defined here forlarge geographical units (countries) that include many differ-ent habitats and the biogeographic patterns defined on sobroad a scale are best characterised by macroenvironmen-tal factors that may go beyond the local habitat needs Abiotic element usually includes species that do not sharehabitats or way of life For example chorotype 2 includesTockus flavirostris and Tockus deckeni which are morecommon in open thorn and semi-arid savanna especially inCommiphora woodland but also includes Tockushemprichii that occupies rocky habitat in hills and gorgesDifferent ecological characteristics of habitat structure (suchas nest site availability or food abundance) may affect thehabitat selection of these species (see for example Hildeacuten1965 Elmberg et al 1994) but the macroenvironmental fac-
tors such as climate could create some basic conditionsthat are shared by the local environments so giving place toa common distribution pattern for species that are ecologi-cally different
Chorotypes also merit some consideration from the pointof view of conservation biology Birks (1987) proposed thatchorotypes could be valuable on the grounds of biogeo-graphical diversity So the degree to which each biotic ele-ment is represented in an area is a useful criterion for eval-uating its conservation interest combined with the richnessthe rarity and the conservation status of the species(Ramiacuterez and Vargas 1992 Turpie and Crowe 1994Williams et al 1996) and for assessing the representative-ness of the area (see Dasmann 1972 Austin and Margules1986 Margules 1986 Saeligtersdal and Birks 1993) Thiswould be useful for hornbill conservation planning on theAfrican scale although the selection of more specific sitesas particular reserves should require a more local and eco-logical approach
Particularly important in this respect are the areas withthe higher number of species of each chorotype which canbe seen in Figures 2 3 and 4 These may be considered asthe geographic cores of the chorotypes and may be spe-cially relevant areas for hornbills Even some disjoint distri-butions affect these areas For example Bycanistes breviswhich displays a patchy distribution due to the patchiness ofthe forest it occupies belongs to chorotype 3 but its north-ern distribution area patch is located in the core of chorotype2 (see Figures 3 and 4)
Macro-environmental and biotic characterisation of thechorotypesThe macro-environmental and biotic analyses of the hornbillchorotypes suggest a possible relationship of sexual dimor-phism with climatic seasonality and habitat type Hornbillsinhabiting non-seasonal areas supporting forest habitats(chorotype 1) have low sexual dimorphism while hornbillinhabiting seasonal savanna and grassland areas(chorotype 2) show greater levels of sexual dimorphismmales being bigger than females Bigger males might befavourably selected when competing for territories in thesavanna where suitable nesting sites are scarcer than in theforest
The species of chorotype 3 have a preference for wood-land or montane and coastal forest but not for rainforestand occur in the countries with colder winters and moredesert scrub and desert forest This could suggest that horn-bills can inhabit desert environments as long as there issome tree or scrub coverage although because of the limi-tation of the scale of this study more specific observationsof habitat usage are needed to assess this possibility
The ungrouped pattern of distributionThe three species that are not grouped in any chorotype arethose with the widest distribution in Africa specially theAfrican Grey Hornbill Tockus nasutus and the Red-billedHornbill Tockus erythrorhynchus several subspecies ofwhich can be found on both sides of the belt of tallBrachystegia or miombo woodland that stretches acrosscentral Africa (Kemp 2001) These two latter species are
Figure 4 Distribution of chorotype 3 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas46
also singled out because of their movements T nasutus hasbeen reported to make regular movements near the Saharaand irregular irruptive movements in southern Africa (Kemp2001) while T erythrorhynchus appears to undertake regu-lar movements all over its distribution area In contrast theNorthern Ground-hornbill Bucorvus abyssinicus resides inthe sub-Saharan African savannas north of the equator(Kemp 1995)
These three species are the only ones occurring inMauritania Mali Niger and Burkina Faso which representthe northernmost distribution for the hornbills in Africa andT nasutus is the only hornbill that is also present in theArabian Peninsula It is note-worthy that if only the Africandistribution of the African Grey Hornbill is considered thenthese three species are grouped together constituting afourth chorotype and then all African hornbill species followa discrete distribution pattern in a strictly African contextHowever the presence of Tockus nasutus in south-east Asiaseems to have somehow disrupted the well established dis-crete biogeographical configuration of the African hornbills
Acknowledgements mdash We are grateful to Dr Alan Kemp Dr PilaiPoonswad Dr Helen de Klerk and Dr Juan Carranza for their criti-cal comments on an early version of this paper This study was fund-ed by the European Commision and Spanish Government project1FD97ndash1571ndashC02ndash01
References
Austin MP and Margules CR 1986 Assessing representativenessIn Usher MB (ed) Wildlife Conservation Evaluation pp 45ndash68Chapman and Hall London
Austin MP and Smith TM 1989 A new model for the continuumconcept Vegetatio 83 35ndash47
Baroni-Urbani C and Buser MW 1976 Similarity of binary dataSystematic Zoology 25 251ndash259
Baroni-Urbani C Rufo S and Vigna-Taglianti A 1978 Materialiper una biogeografia italiana fondata su alguni generi dicoleotteri cicindelidi carabidi e crisomelidi Estratto dellaMemorie della Societa Entomologica Italiana 56 35ndash92
Birks HJB 1987 Recent methodological development in quantita-tive descriptive biogeography Annales Zoology Fennici 24165ndash178
Brown JH 1995 Macroecology The University of Chicago PressChicago
Brown JH 1999 Macroecology progress and prospect Oikos 873ndash14
Cordellier S and Didiot B 1999 Lrsquoeacutetet du monde Eacuteditions LaDeacutecouverte et Syros Pariacutes
Crowe TM and Crowe AA 1982 Patterns in distribution diversityand endemism in Afrotropical birds Journal of Zoology London198 417ndash442
Crowe TM and Kemp AC 1988 African historical biogeography asreflected by Galliform and Hornbill evolution In Ouellet H (ed)Acta XIX Congressus Internationalis Ornithologici Vol 2 pp2510ndash2518 University of Ottawa Press Ottawa
Dasmann RF 1972 Towards a system for classifying naturalregions of the world and their representation by National Parksand Reserves Biological Conservation 4 247ndash255
Elmberg J Sjoumlberg K Nummi P and Poumlysauml H 1994 Patterns oflake acidity and waterfowl communities Hydrobiologia 279280201ndash206
FAO 1984 Agroclimatological data for Africa FAO Plant Productionand Protection Series No 22 Vols 1 and 2 Rome
Guillet A and Crowe TM 1985 Patterns of distribution speciesrichness endemism and guild composition of water-birds inAfrica African Journal of Ecology 23 89ndash120
Guillet A and Crowe TM 1986 A preliminary investigation of pat-terns of distribution and species richness of southern Africanwaterbirds South African Journal of Wildlife Research 1665ndash81
Hengeveld R 1990 Dynamic Biogeography Cambridge UniversityPress Cambridge
Hengeveld R and Hogeweg P 1979 Cluster analysis of the distri-bution patterns of Dutch carabid species (Col) In Orloci L RaoCR and Stiteler WM (eds) Multivariate Methods in EcologicalWork pp 65ndash86 International Co-operative Publishing HouseBurtonsville
Hildeacuten O 1965 Habitat selection in birds a review AnnalesZoology Fennici 2 53ndash75
Hosmer DW and Lemeshow S 1989 Applied Logistic RegressionJohn Wiley and Sons Inc New York
Kemp A 1995 The Hornbills Bucerotiformes Oxford UniversityPress
Kemp A 2001 Family Bucerotidae (Hornbills) In Del Hoyo J ElliotA and Sargatal J (eds) Handbook of the Birds of the World Vol6 Mousebirds to Hornbills pp 436ndash523 Lynx Editions Barcelona
Kikkawa J and Pearse K 1969 Geographical distribution of landbirds in Australia a numerical analysis Australian Journal ofZoology 17 821ndash840
Lance GN and Williams WT 1967 A general theory of classificato-ry sorting strategies I Hierarchical systems Computer Journa l9 373ndash380
Margules CR 1986 Conservation evaluation in practice In UsherMB (ed) Wildlife Conservation Evaluation pp 298ndash314 Chapmanand Hall London
Maacuterquez AL Real R Vargas JM and Salvo AE 1997 On identify-ing common distribution patterns and their causal factors a prob-abilistic method applied to pteridophytes in the Iberian PeninsulaJournal of Biogeography 24 613ndash631
Mayr E 1965 What is a fauna Zoologisches Jahrbuch derSystematik 92 473ndash486
McCoy ED Bell SS and Walters K 1986 Identifying bioticboundaries along environmental gradients Ecology 67 749ndash759
Olivero J Real R and Vargas JM 1998 Distribution of breedingwintering and resident waterbirds in Europe biotic regions andthe macroclimate Ornis Fennica 75 153ndash175
Rabinovich JE and Rapoport EH 1975 Geographical variation ofdiversity in passerine birds Journal of Biogeography 2141ndash157
Ramiacuterez JM and Vargas JM 1992 Contribucioacuten de la biogeografiacuteaa la gestioacuten del medio ambiente y a la conservacioacuten de lasespecies In Vargas JM Real R and Antuacutenez A (eds) Objetivos yMeacutetodos Biogeograacuteficos Aplicaciones en Herpetologiacutea 2 pp95ndash106 Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Vargas JM and Guerrero JC 1992a Anaacutelisis biogeograacuteficode clasificacioacuten de aacutereas y especies In Vargas JM Real R andAntuacutenez A (ed) Objetivos y Meacutetodos Biogeograacuteficos Aplicacionesen Herpetologiacutea Monografias Herpetoloacutegica 2 pp 73ndash84Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Guerrero JC and Ramiacuterez JM 1992b Identificacioacuten defronteras bioacuteticas significativas para los anfibios en la cuencahidrograacutefica del Sur de Espantildea Dontildeana Acta Vertebrata 1953ndash70
Real R Pleguezuelos JM and Fahd S 1997 The distribution pat-terns of reptiles in the Riff region northern Morocco AfricanJournal of Ecology 35 312ndash325
Saeligtersdal M and Birks HJB 1993 Assessing the representative-ness of nature reserves using multivariate analysis vascularplants and breeding birds in deciduous forest western Norway
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 47
Biological Conservation 65 121ndash132Sibley CS and Monroe BE 1990 Distribution and Taxonomy of
Birds of the World Yale University Press New HavenSneath PHA and Sokal RR 1973 Numerical Taxonomy The
Principles and Practices of Numerical Classification FreemanSan Francisco
Sokal RR and Rohlf FJ 1981 Biometry The Principles andPractice of Statistics in Biological Research WH Freeman andCompany New York
Turner JRG Lennon JJ and Lawrenson JA 1988 British birdspecies distributions and the energy theory Nature 355 539ndash541
Turpie J K and Crowe TM 1994 Patterns of distribution diversityand endemism of larger African mammals South AfricanJournal of Zoology 29 19ndash32
Whitehead PJ Bowman DMJS and Tideman SC 1992Biogeographic patterns environmental correlates and conserva-tion of avifauna in the Northern Territory Australia Journal ofBiogeography 19 151ndash161
Wiens JA 1989 Spatial scaling in ecology Functional Ecology 3385ndash397
Williams PH De Klerk HM and Crowe TM 1999 Interpreting bio-geographical boundaries among Afrotropical birds spatial pat-terns in richness gradients and species replacement Journal ofBiogeography 26 459ndash474
Williams P Gibbons D Margules C Rebelo A Humphries C andPressey R 1996 A comparison of richness hotspots rarityhotspots and complementary areas for conserving diversity ofBritish birds Conservation Biology 10 155ndash174
World Resources Institute 1994 World Resources 1994ndash95 DataBase Diskette World Resources Institute Washington DC
Wright GH 1983 Species-energy theory an extension to species-area theory Oikos 41 496ndash506
Resumen
Se ha estudiado el aacuterea de distribucioacuten por paiacuteses de las 23 especies de caacutelaos africanos con el fin de identificar grupos dedistribuciones significativamente similares es decir corotipos y distribuciones que se reemplazan gradualmente Se hautilizado un meacutetodo probabiliacutestico de clasificacioacuten para distinguir entre ambos tipos de patrones biogeograacuteficos Una vezobtenidos los corotipos se analizoacute como se ven afectados por la influencia del clima el aacuterea la superficie cubierta por losdiferentes tipos de haacutebitat y las perturbaciones humanas en cada paiacutes Tambieacuten se comproboacute si las caracteriacutesticas propiasde cada especie entre las que se encuentran variables relacionadas con el haacutebitat la alimentacioacuten su biometriacutea y diferentesparaacutemetros metaboacutelicos influyen en los patrones biogeograacuteficos detectados Se han encontrado 3 corotipos que incluyen a20 especies El corotipo centro-occidental estaacute formado por 9 especies caracterizadas por la preferencia por haacutebitats debosque y por el bajo dimorfismo sexual y por 25 paiacuteses caracterizados por presentar un rango de temperaturas muy bajoEl corotipo oriental incluye a tres especies que se caracterizan por presentar haacutebitos alimenticios terrestres y un elevadodimorfismo sexual y a 7 paiacuteses caracterizados por mostrar gran parte de su superficie cubierta por sabanas y pastos Elcorotipo meridional agrupa la distribucioacuten de 8 especies que se caracterizan por su preferencia por bosques de montantildea ycosteros y zonas arboladas y a 18 paiacuteses con temperaturas bajas en el mes maacutes friacuteo y grandes superficies cubiertas pormatorrales y bosques de desierto Hay 3 especies que no se agrupan en corotipos y que siguen un reemplazamiento graduala lo largo del aacuterea de estudio
Chorotype 1African Pied Hornbill Tockus fasciatusPiping Hornbill Bycanistes fistulatorRed-billed Dwarf-hornbill Tockus camurusBlack-casqued Hornbill Ceratogymna atrataBrown-cheeked Hornbill Bycanistes cylindricusLong-tailed Hornbill Tropicranus albocristatusBlack Dwarf-hornbill Tockus hartlaubiGrey-cheeked Hornbill Bycanistes subcylindricusYellow-casqued Hornbill Ceratogymna elata
Chorotype 2Eastern Yellow-billed Hornbill Tockus flavirostrisVon der Deckenrsquos Hornbill Tockus deckeniHemprichrsquos Hornbill Tockus hemprichii
Appendix 1 Biotic elements of the three chorotypes obtained in Africa for the hornbills The corresponding geographic elements are shownin Figures 2 3 and 4 The species nomenclature follows Kemp (2001)
Chorotype 3Southern Ground-hornbill Bucorvus leadbeateriCrowned Hornbill Tockus alboterminatusPale-billed Hornbill Tockus pallidirostrisTrumpeter Hornbill Bycanistes bucinatorSilvery-cheeked Hornbill Bycanistes brevisBradfieldrsquos Hornbill Tockus bradfieldiSouthern Yellow-billed Hornbill Tockus leucomelasMonteirorsquos Hornbill Tockus monteiri
Ungrouped speciesAfrican Grey Hornbill Tockus nasutusRed Billed Hornbill Tockus erythrorhynchusNorthern Ground-hornbill Bucorvus abyssinicus
Received August 2001 accepted July 2002Editor WRJ Dean
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas44
The geographic element is dominated by desert scruband desert forest and is characterised by a low temperaturein the coldest monthy = 382747 ndash17272 (CT) +00140 (DesertScrub) +00642 (DesertF)
All the segregations between chorotypes were strongThe strongest segregation occurs between chorotypes 1and 3 (DS = 06686 GS = 1406738 P lt 0001) followed bythe segregation between chorotypes 1 and 2 (DS = 06381GS = 586733 P lt 0001) and between chorotypes 2 and 3(DS = 03683 GS = 141756 P lt 0001)
The distributions of the three species that are not includ-ed in any chorotype are more similar to chorotype 1 than tothe other chorotypes (Figure 1) Two of them have a widedistribution throughout the continent the African GreyHornbill Tockus nasutus and the Red-billed Hornbill Tockuserythrorhynchus The other species is the Northern Ground-hornbill Bucorvus abyssinicus which occupies the sub-Saharan fringe from west to east
Discussion
The meaning of chorotypesThe majority of hornbill species in Africa (20 of the 23species) exhibit a well established discrete biogeographical
pattern constituted by three chorotypes with almost no over-lap occupying the western-central the eastern and thesouthern countries of Africa respectively Real et al (1997)considered that the fact that a large percentage of speciesconstitute chorotypes may be indicative of biogeographicalequilibrium in the biota analysed that is that species distri-butions have been well established for a long timeAccording to this interpretation hornbills would be a group ofspecies with long established distribution ranges in Africa
As a superspecies pair is shared between chorotypestwo and three in the form of Tockus flavirostris and T leu-
Figure 1 Dendrogram of classification of the distributions of Africanhornbills where the three significant chorotypes are signalled
Table 5 Segregations between groups of species on the dendro-gram nodes of Figure 1 The species of each chorotype and of theungrouped species appear in Appendix 1 DW(AxA) quantify theinternal homogeneity of each group GW(AxA) is a statisticalparameter obtained using a G test of independence ns = non sig-nificant Critical values for GW(AxA) are those of the Chi-squaredistribution with one degree of freedom
Group Cophenetic DW(AxA) GW(AxA) Significancecoefficient
Chorotype 1 0682 0557 47435 P lt 0001Chorotype 2 0870 0627 7227 P lt 001Chorotype 3 0560 0426 16429 P lt 0001Ungrouped species 0637 0415 1606 ns
Figure 2 Distribution of chorotype 1 with the number of species ofthe chorotype overlapping in each area
Figure 3 Distribution of chorotype 2 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 45
comelas these two chorotypes may be considered to bemore closely related to each other than to chorotype 1 asFigure 1 also suggests
The three chorotypes obtained are almost coincidentwith the Forest lsquoHornrsquo of Africa and Southern Savanna sub-regions respectively obtained by Williams et al (1999) for1911 species of Afrotropical birds (see also Crowe andCrowe 1982) This may indicate that this regionalisation notonly reflects the spatial responses of hornbills to environ-mental conditions but that these environmental factorsaffect the distributions of most Afrotropical birds at a conti-nental scale
The biotic element of the chorotypes should be distin-guished from ecological assemblages which imply the shar-ing of habitat or guilds which use the same trophicresources Chorotypes should rather be seen in a macroe-cological perspective (Brown 1995 Brown 1999) as emer-gent biogeographical structures generated by complex eco-logical systems which assemblages and guilds could beparticulate components of Chorotypes are defined here forlarge geographical units (countries) that include many differ-ent habitats and the biogeographic patterns defined on sobroad a scale are best characterised by macroenvironmen-tal factors that may go beyond the local habitat needs Abiotic element usually includes species that do not sharehabitats or way of life For example chorotype 2 includesTockus flavirostris and Tockus deckeni which are morecommon in open thorn and semi-arid savanna especially inCommiphora woodland but also includes Tockushemprichii that occupies rocky habitat in hills and gorgesDifferent ecological characteristics of habitat structure (suchas nest site availability or food abundance) may affect thehabitat selection of these species (see for example Hildeacuten1965 Elmberg et al 1994) but the macroenvironmental fac-
tors such as climate could create some basic conditionsthat are shared by the local environments so giving place toa common distribution pattern for species that are ecologi-cally different
Chorotypes also merit some consideration from the pointof view of conservation biology Birks (1987) proposed thatchorotypes could be valuable on the grounds of biogeo-graphical diversity So the degree to which each biotic ele-ment is represented in an area is a useful criterion for eval-uating its conservation interest combined with the richnessthe rarity and the conservation status of the species(Ramiacuterez and Vargas 1992 Turpie and Crowe 1994Williams et al 1996) and for assessing the representative-ness of the area (see Dasmann 1972 Austin and Margules1986 Margules 1986 Saeligtersdal and Birks 1993) Thiswould be useful for hornbill conservation planning on theAfrican scale although the selection of more specific sitesas particular reserves should require a more local and eco-logical approach
Particularly important in this respect are the areas withthe higher number of species of each chorotype which canbe seen in Figures 2 3 and 4 These may be considered asthe geographic cores of the chorotypes and may be spe-cially relevant areas for hornbills Even some disjoint distri-butions affect these areas For example Bycanistes breviswhich displays a patchy distribution due to the patchiness ofthe forest it occupies belongs to chorotype 3 but its north-ern distribution area patch is located in the core of chorotype2 (see Figures 3 and 4)
Macro-environmental and biotic characterisation of thechorotypesThe macro-environmental and biotic analyses of the hornbillchorotypes suggest a possible relationship of sexual dimor-phism with climatic seasonality and habitat type Hornbillsinhabiting non-seasonal areas supporting forest habitats(chorotype 1) have low sexual dimorphism while hornbillinhabiting seasonal savanna and grassland areas(chorotype 2) show greater levels of sexual dimorphismmales being bigger than females Bigger males might befavourably selected when competing for territories in thesavanna where suitable nesting sites are scarcer than in theforest
The species of chorotype 3 have a preference for wood-land or montane and coastal forest but not for rainforestand occur in the countries with colder winters and moredesert scrub and desert forest This could suggest that horn-bills can inhabit desert environments as long as there issome tree or scrub coverage although because of the limi-tation of the scale of this study more specific observationsof habitat usage are needed to assess this possibility
The ungrouped pattern of distributionThe three species that are not grouped in any chorotype arethose with the widest distribution in Africa specially theAfrican Grey Hornbill Tockus nasutus and the Red-billedHornbill Tockus erythrorhynchus several subspecies ofwhich can be found on both sides of the belt of tallBrachystegia or miombo woodland that stretches acrosscentral Africa (Kemp 2001) These two latter species are
Figure 4 Distribution of chorotype 3 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas46
also singled out because of their movements T nasutus hasbeen reported to make regular movements near the Saharaand irregular irruptive movements in southern Africa (Kemp2001) while T erythrorhynchus appears to undertake regu-lar movements all over its distribution area In contrast theNorthern Ground-hornbill Bucorvus abyssinicus resides inthe sub-Saharan African savannas north of the equator(Kemp 1995)
These three species are the only ones occurring inMauritania Mali Niger and Burkina Faso which representthe northernmost distribution for the hornbills in Africa andT nasutus is the only hornbill that is also present in theArabian Peninsula It is note-worthy that if only the Africandistribution of the African Grey Hornbill is considered thenthese three species are grouped together constituting afourth chorotype and then all African hornbill species followa discrete distribution pattern in a strictly African contextHowever the presence of Tockus nasutus in south-east Asiaseems to have somehow disrupted the well established dis-crete biogeographical configuration of the African hornbills
Acknowledgements mdash We are grateful to Dr Alan Kemp Dr PilaiPoonswad Dr Helen de Klerk and Dr Juan Carranza for their criti-cal comments on an early version of this paper This study was fund-ed by the European Commision and Spanish Government project1FD97ndash1571ndashC02ndash01
References
Austin MP and Margules CR 1986 Assessing representativenessIn Usher MB (ed) Wildlife Conservation Evaluation pp 45ndash68Chapman and Hall London
Austin MP and Smith TM 1989 A new model for the continuumconcept Vegetatio 83 35ndash47
Baroni-Urbani C and Buser MW 1976 Similarity of binary dataSystematic Zoology 25 251ndash259
Baroni-Urbani C Rufo S and Vigna-Taglianti A 1978 Materialiper una biogeografia italiana fondata su alguni generi dicoleotteri cicindelidi carabidi e crisomelidi Estratto dellaMemorie della Societa Entomologica Italiana 56 35ndash92
Birks HJB 1987 Recent methodological development in quantita-tive descriptive biogeography Annales Zoology Fennici 24165ndash178
Brown JH 1995 Macroecology The University of Chicago PressChicago
Brown JH 1999 Macroecology progress and prospect Oikos 873ndash14
Cordellier S and Didiot B 1999 Lrsquoeacutetet du monde Eacuteditions LaDeacutecouverte et Syros Pariacutes
Crowe TM and Crowe AA 1982 Patterns in distribution diversityand endemism in Afrotropical birds Journal of Zoology London198 417ndash442
Crowe TM and Kemp AC 1988 African historical biogeography asreflected by Galliform and Hornbill evolution In Ouellet H (ed)Acta XIX Congressus Internationalis Ornithologici Vol 2 pp2510ndash2518 University of Ottawa Press Ottawa
Dasmann RF 1972 Towards a system for classifying naturalregions of the world and their representation by National Parksand Reserves Biological Conservation 4 247ndash255
Elmberg J Sjoumlberg K Nummi P and Poumlysauml H 1994 Patterns oflake acidity and waterfowl communities Hydrobiologia 279280201ndash206
FAO 1984 Agroclimatological data for Africa FAO Plant Productionand Protection Series No 22 Vols 1 and 2 Rome
Guillet A and Crowe TM 1985 Patterns of distribution speciesrichness endemism and guild composition of water-birds inAfrica African Journal of Ecology 23 89ndash120
Guillet A and Crowe TM 1986 A preliminary investigation of pat-terns of distribution and species richness of southern Africanwaterbirds South African Journal of Wildlife Research 1665ndash81
Hengeveld R 1990 Dynamic Biogeography Cambridge UniversityPress Cambridge
Hengeveld R and Hogeweg P 1979 Cluster analysis of the distri-bution patterns of Dutch carabid species (Col) In Orloci L RaoCR and Stiteler WM (eds) Multivariate Methods in EcologicalWork pp 65ndash86 International Co-operative Publishing HouseBurtonsville
Hildeacuten O 1965 Habitat selection in birds a review AnnalesZoology Fennici 2 53ndash75
Hosmer DW and Lemeshow S 1989 Applied Logistic RegressionJohn Wiley and Sons Inc New York
Kemp A 1995 The Hornbills Bucerotiformes Oxford UniversityPress
Kemp A 2001 Family Bucerotidae (Hornbills) In Del Hoyo J ElliotA and Sargatal J (eds) Handbook of the Birds of the World Vol6 Mousebirds to Hornbills pp 436ndash523 Lynx Editions Barcelona
Kikkawa J and Pearse K 1969 Geographical distribution of landbirds in Australia a numerical analysis Australian Journal ofZoology 17 821ndash840
Lance GN and Williams WT 1967 A general theory of classificato-ry sorting strategies I Hierarchical systems Computer Journa l9 373ndash380
Margules CR 1986 Conservation evaluation in practice In UsherMB (ed) Wildlife Conservation Evaluation pp 298ndash314 Chapmanand Hall London
Maacuterquez AL Real R Vargas JM and Salvo AE 1997 On identify-ing common distribution patterns and their causal factors a prob-abilistic method applied to pteridophytes in the Iberian PeninsulaJournal of Biogeography 24 613ndash631
Mayr E 1965 What is a fauna Zoologisches Jahrbuch derSystematik 92 473ndash486
McCoy ED Bell SS and Walters K 1986 Identifying bioticboundaries along environmental gradients Ecology 67 749ndash759
Olivero J Real R and Vargas JM 1998 Distribution of breedingwintering and resident waterbirds in Europe biotic regions andthe macroclimate Ornis Fennica 75 153ndash175
Rabinovich JE and Rapoport EH 1975 Geographical variation ofdiversity in passerine birds Journal of Biogeography 2141ndash157
Ramiacuterez JM and Vargas JM 1992 Contribucioacuten de la biogeografiacuteaa la gestioacuten del medio ambiente y a la conservacioacuten de lasespecies In Vargas JM Real R and Antuacutenez A (eds) Objetivos yMeacutetodos Biogeograacuteficos Aplicaciones en Herpetologiacutea 2 pp95ndash106 Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Vargas JM and Guerrero JC 1992a Anaacutelisis biogeograacuteficode clasificacioacuten de aacutereas y especies In Vargas JM Real R andAntuacutenez A (ed) Objetivos y Meacutetodos Biogeograacuteficos Aplicacionesen Herpetologiacutea Monografias Herpetoloacutegica 2 pp 73ndash84Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Guerrero JC and Ramiacuterez JM 1992b Identificacioacuten defronteras bioacuteticas significativas para los anfibios en la cuencahidrograacutefica del Sur de Espantildea Dontildeana Acta Vertebrata 1953ndash70
Real R Pleguezuelos JM and Fahd S 1997 The distribution pat-terns of reptiles in the Riff region northern Morocco AfricanJournal of Ecology 35 312ndash325
Saeligtersdal M and Birks HJB 1993 Assessing the representative-ness of nature reserves using multivariate analysis vascularplants and breeding birds in deciduous forest western Norway
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 47
Biological Conservation 65 121ndash132Sibley CS and Monroe BE 1990 Distribution and Taxonomy of
Birds of the World Yale University Press New HavenSneath PHA and Sokal RR 1973 Numerical Taxonomy The
Principles and Practices of Numerical Classification FreemanSan Francisco
Sokal RR and Rohlf FJ 1981 Biometry The Principles andPractice of Statistics in Biological Research WH Freeman andCompany New York
Turner JRG Lennon JJ and Lawrenson JA 1988 British birdspecies distributions and the energy theory Nature 355 539ndash541
Turpie J K and Crowe TM 1994 Patterns of distribution diversityand endemism of larger African mammals South AfricanJournal of Zoology 29 19ndash32
Whitehead PJ Bowman DMJS and Tideman SC 1992Biogeographic patterns environmental correlates and conserva-tion of avifauna in the Northern Territory Australia Journal ofBiogeography 19 151ndash161
Wiens JA 1989 Spatial scaling in ecology Functional Ecology 3385ndash397
Williams PH De Klerk HM and Crowe TM 1999 Interpreting bio-geographical boundaries among Afrotropical birds spatial pat-terns in richness gradients and species replacement Journal ofBiogeography 26 459ndash474
Williams P Gibbons D Margules C Rebelo A Humphries C andPressey R 1996 A comparison of richness hotspots rarityhotspots and complementary areas for conserving diversity ofBritish birds Conservation Biology 10 155ndash174
World Resources Institute 1994 World Resources 1994ndash95 DataBase Diskette World Resources Institute Washington DC
Wright GH 1983 Species-energy theory an extension to species-area theory Oikos 41 496ndash506
Resumen
Se ha estudiado el aacuterea de distribucioacuten por paiacuteses de las 23 especies de caacutelaos africanos con el fin de identificar grupos dedistribuciones significativamente similares es decir corotipos y distribuciones que se reemplazan gradualmente Se hautilizado un meacutetodo probabiliacutestico de clasificacioacuten para distinguir entre ambos tipos de patrones biogeograacuteficos Una vezobtenidos los corotipos se analizoacute como se ven afectados por la influencia del clima el aacuterea la superficie cubierta por losdiferentes tipos de haacutebitat y las perturbaciones humanas en cada paiacutes Tambieacuten se comproboacute si las caracteriacutesticas propiasde cada especie entre las que se encuentran variables relacionadas con el haacutebitat la alimentacioacuten su biometriacutea y diferentesparaacutemetros metaboacutelicos influyen en los patrones biogeograacuteficos detectados Se han encontrado 3 corotipos que incluyen a20 especies El corotipo centro-occidental estaacute formado por 9 especies caracterizadas por la preferencia por haacutebitats debosque y por el bajo dimorfismo sexual y por 25 paiacuteses caracterizados por presentar un rango de temperaturas muy bajoEl corotipo oriental incluye a tres especies que se caracterizan por presentar haacutebitos alimenticios terrestres y un elevadodimorfismo sexual y a 7 paiacuteses caracterizados por mostrar gran parte de su superficie cubierta por sabanas y pastos Elcorotipo meridional agrupa la distribucioacuten de 8 especies que se caracterizan por su preferencia por bosques de montantildea ycosteros y zonas arboladas y a 18 paiacuteses con temperaturas bajas en el mes maacutes friacuteo y grandes superficies cubiertas pormatorrales y bosques de desierto Hay 3 especies que no se agrupan en corotipos y que siguen un reemplazamiento graduala lo largo del aacuterea de estudio
Chorotype 1African Pied Hornbill Tockus fasciatusPiping Hornbill Bycanistes fistulatorRed-billed Dwarf-hornbill Tockus camurusBlack-casqued Hornbill Ceratogymna atrataBrown-cheeked Hornbill Bycanistes cylindricusLong-tailed Hornbill Tropicranus albocristatusBlack Dwarf-hornbill Tockus hartlaubiGrey-cheeked Hornbill Bycanistes subcylindricusYellow-casqued Hornbill Ceratogymna elata
Chorotype 2Eastern Yellow-billed Hornbill Tockus flavirostrisVon der Deckenrsquos Hornbill Tockus deckeniHemprichrsquos Hornbill Tockus hemprichii
Appendix 1 Biotic elements of the three chorotypes obtained in Africa for the hornbills The corresponding geographic elements are shownin Figures 2 3 and 4 The species nomenclature follows Kemp (2001)
Chorotype 3Southern Ground-hornbill Bucorvus leadbeateriCrowned Hornbill Tockus alboterminatusPale-billed Hornbill Tockus pallidirostrisTrumpeter Hornbill Bycanistes bucinatorSilvery-cheeked Hornbill Bycanistes brevisBradfieldrsquos Hornbill Tockus bradfieldiSouthern Yellow-billed Hornbill Tockus leucomelasMonteirorsquos Hornbill Tockus monteiri
Ungrouped speciesAfrican Grey Hornbill Tockus nasutusRed Billed Hornbill Tockus erythrorhynchusNorthern Ground-hornbill Bucorvus abyssinicus
Received August 2001 accepted July 2002Editor WRJ Dean
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 45
comelas these two chorotypes may be considered to bemore closely related to each other than to chorotype 1 asFigure 1 also suggests
The three chorotypes obtained are almost coincidentwith the Forest lsquoHornrsquo of Africa and Southern Savanna sub-regions respectively obtained by Williams et al (1999) for1911 species of Afrotropical birds (see also Crowe andCrowe 1982) This may indicate that this regionalisation notonly reflects the spatial responses of hornbills to environ-mental conditions but that these environmental factorsaffect the distributions of most Afrotropical birds at a conti-nental scale
The biotic element of the chorotypes should be distin-guished from ecological assemblages which imply the shar-ing of habitat or guilds which use the same trophicresources Chorotypes should rather be seen in a macroe-cological perspective (Brown 1995 Brown 1999) as emer-gent biogeographical structures generated by complex eco-logical systems which assemblages and guilds could beparticulate components of Chorotypes are defined here forlarge geographical units (countries) that include many differ-ent habitats and the biogeographic patterns defined on sobroad a scale are best characterised by macroenvironmen-tal factors that may go beyond the local habitat needs Abiotic element usually includes species that do not sharehabitats or way of life For example chorotype 2 includesTockus flavirostris and Tockus deckeni which are morecommon in open thorn and semi-arid savanna especially inCommiphora woodland but also includes Tockushemprichii that occupies rocky habitat in hills and gorgesDifferent ecological characteristics of habitat structure (suchas nest site availability or food abundance) may affect thehabitat selection of these species (see for example Hildeacuten1965 Elmberg et al 1994) but the macroenvironmental fac-
tors such as climate could create some basic conditionsthat are shared by the local environments so giving place toa common distribution pattern for species that are ecologi-cally different
Chorotypes also merit some consideration from the pointof view of conservation biology Birks (1987) proposed thatchorotypes could be valuable on the grounds of biogeo-graphical diversity So the degree to which each biotic ele-ment is represented in an area is a useful criterion for eval-uating its conservation interest combined with the richnessthe rarity and the conservation status of the species(Ramiacuterez and Vargas 1992 Turpie and Crowe 1994Williams et al 1996) and for assessing the representative-ness of the area (see Dasmann 1972 Austin and Margules1986 Margules 1986 Saeligtersdal and Birks 1993) Thiswould be useful for hornbill conservation planning on theAfrican scale although the selection of more specific sitesas particular reserves should require a more local and eco-logical approach
Particularly important in this respect are the areas withthe higher number of species of each chorotype which canbe seen in Figures 2 3 and 4 These may be considered asthe geographic cores of the chorotypes and may be spe-cially relevant areas for hornbills Even some disjoint distri-butions affect these areas For example Bycanistes breviswhich displays a patchy distribution due to the patchiness ofthe forest it occupies belongs to chorotype 3 but its north-ern distribution area patch is located in the core of chorotype2 (see Figures 3 and 4)
Macro-environmental and biotic characterisation of thechorotypesThe macro-environmental and biotic analyses of the hornbillchorotypes suggest a possible relationship of sexual dimor-phism with climatic seasonality and habitat type Hornbillsinhabiting non-seasonal areas supporting forest habitats(chorotype 1) have low sexual dimorphism while hornbillinhabiting seasonal savanna and grassland areas(chorotype 2) show greater levels of sexual dimorphismmales being bigger than females Bigger males might befavourably selected when competing for territories in thesavanna where suitable nesting sites are scarcer than in theforest
The species of chorotype 3 have a preference for wood-land or montane and coastal forest but not for rainforestand occur in the countries with colder winters and moredesert scrub and desert forest This could suggest that horn-bills can inhabit desert environments as long as there issome tree or scrub coverage although because of the limi-tation of the scale of this study more specific observationsof habitat usage are needed to assess this possibility
The ungrouped pattern of distributionThe three species that are not grouped in any chorotype arethose with the widest distribution in Africa specially theAfrican Grey Hornbill Tockus nasutus and the Red-billedHornbill Tockus erythrorhynchus several subspecies ofwhich can be found on both sides of the belt of tallBrachystegia or miombo woodland that stretches acrosscentral Africa (Kemp 2001) These two latter species are
Figure 4 Distribution of chorotype 3 with the number of species ofthe chorotype overlapping in each area
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas46
also singled out because of their movements T nasutus hasbeen reported to make regular movements near the Saharaand irregular irruptive movements in southern Africa (Kemp2001) while T erythrorhynchus appears to undertake regu-lar movements all over its distribution area In contrast theNorthern Ground-hornbill Bucorvus abyssinicus resides inthe sub-Saharan African savannas north of the equator(Kemp 1995)
These three species are the only ones occurring inMauritania Mali Niger and Burkina Faso which representthe northernmost distribution for the hornbills in Africa andT nasutus is the only hornbill that is also present in theArabian Peninsula It is note-worthy that if only the Africandistribution of the African Grey Hornbill is considered thenthese three species are grouped together constituting afourth chorotype and then all African hornbill species followa discrete distribution pattern in a strictly African contextHowever the presence of Tockus nasutus in south-east Asiaseems to have somehow disrupted the well established dis-crete biogeographical configuration of the African hornbills
Acknowledgements mdash We are grateful to Dr Alan Kemp Dr PilaiPoonswad Dr Helen de Klerk and Dr Juan Carranza for their criti-cal comments on an early version of this paper This study was fund-ed by the European Commision and Spanish Government project1FD97ndash1571ndashC02ndash01
References
Austin MP and Margules CR 1986 Assessing representativenessIn Usher MB (ed) Wildlife Conservation Evaluation pp 45ndash68Chapman and Hall London
Austin MP and Smith TM 1989 A new model for the continuumconcept Vegetatio 83 35ndash47
Baroni-Urbani C and Buser MW 1976 Similarity of binary dataSystematic Zoology 25 251ndash259
Baroni-Urbani C Rufo S and Vigna-Taglianti A 1978 Materialiper una biogeografia italiana fondata su alguni generi dicoleotteri cicindelidi carabidi e crisomelidi Estratto dellaMemorie della Societa Entomologica Italiana 56 35ndash92
Birks HJB 1987 Recent methodological development in quantita-tive descriptive biogeography Annales Zoology Fennici 24165ndash178
Brown JH 1995 Macroecology The University of Chicago PressChicago
Brown JH 1999 Macroecology progress and prospect Oikos 873ndash14
Cordellier S and Didiot B 1999 Lrsquoeacutetet du monde Eacuteditions LaDeacutecouverte et Syros Pariacutes
Crowe TM and Crowe AA 1982 Patterns in distribution diversityand endemism in Afrotropical birds Journal of Zoology London198 417ndash442
Crowe TM and Kemp AC 1988 African historical biogeography asreflected by Galliform and Hornbill evolution In Ouellet H (ed)Acta XIX Congressus Internationalis Ornithologici Vol 2 pp2510ndash2518 University of Ottawa Press Ottawa
Dasmann RF 1972 Towards a system for classifying naturalregions of the world and their representation by National Parksand Reserves Biological Conservation 4 247ndash255
Elmberg J Sjoumlberg K Nummi P and Poumlysauml H 1994 Patterns oflake acidity and waterfowl communities Hydrobiologia 279280201ndash206
FAO 1984 Agroclimatological data for Africa FAO Plant Productionand Protection Series No 22 Vols 1 and 2 Rome
Guillet A and Crowe TM 1985 Patterns of distribution speciesrichness endemism and guild composition of water-birds inAfrica African Journal of Ecology 23 89ndash120
Guillet A and Crowe TM 1986 A preliminary investigation of pat-terns of distribution and species richness of southern Africanwaterbirds South African Journal of Wildlife Research 1665ndash81
Hengeveld R 1990 Dynamic Biogeography Cambridge UniversityPress Cambridge
Hengeveld R and Hogeweg P 1979 Cluster analysis of the distri-bution patterns of Dutch carabid species (Col) In Orloci L RaoCR and Stiteler WM (eds) Multivariate Methods in EcologicalWork pp 65ndash86 International Co-operative Publishing HouseBurtonsville
Hildeacuten O 1965 Habitat selection in birds a review AnnalesZoology Fennici 2 53ndash75
Hosmer DW and Lemeshow S 1989 Applied Logistic RegressionJohn Wiley and Sons Inc New York
Kemp A 1995 The Hornbills Bucerotiformes Oxford UniversityPress
Kemp A 2001 Family Bucerotidae (Hornbills) In Del Hoyo J ElliotA and Sargatal J (eds) Handbook of the Birds of the World Vol6 Mousebirds to Hornbills pp 436ndash523 Lynx Editions Barcelona
Kikkawa J and Pearse K 1969 Geographical distribution of landbirds in Australia a numerical analysis Australian Journal ofZoology 17 821ndash840
Lance GN and Williams WT 1967 A general theory of classificato-ry sorting strategies I Hierarchical systems Computer Journa l9 373ndash380
Margules CR 1986 Conservation evaluation in practice In UsherMB (ed) Wildlife Conservation Evaluation pp 298ndash314 Chapmanand Hall London
Maacuterquez AL Real R Vargas JM and Salvo AE 1997 On identify-ing common distribution patterns and their causal factors a prob-abilistic method applied to pteridophytes in the Iberian PeninsulaJournal of Biogeography 24 613ndash631
Mayr E 1965 What is a fauna Zoologisches Jahrbuch derSystematik 92 473ndash486
McCoy ED Bell SS and Walters K 1986 Identifying bioticboundaries along environmental gradients Ecology 67 749ndash759
Olivero J Real R and Vargas JM 1998 Distribution of breedingwintering and resident waterbirds in Europe biotic regions andthe macroclimate Ornis Fennica 75 153ndash175
Rabinovich JE and Rapoport EH 1975 Geographical variation ofdiversity in passerine birds Journal of Biogeography 2141ndash157
Ramiacuterez JM and Vargas JM 1992 Contribucioacuten de la biogeografiacuteaa la gestioacuten del medio ambiente y a la conservacioacuten de lasespecies In Vargas JM Real R and Antuacutenez A (eds) Objetivos yMeacutetodos Biogeograacuteficos Aplicaciones en Herpetologiacutea 2 pp95ndash106 Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Vargas JM and Guerrero JC 1992a Anaacutelisis biogeograacuteficode clasificacioacuten de aacutereas y especies In Vargas JM Real R andAntuacutenez A (ed) Objetivos y Meacutetodos Biogeograacuteficos Aplicacionesen Herpetologiacutea Monografias Herpetoloacutegica 2 pp 73ndash84Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Guerrero JC and Ramiacuterez JM 1992b Identificacioacuten defronteras bioacuteticas significativas para los anfibios en la cuencahidrograacutefica del Sur de Espantildea Dontildeana Acta Vertebrata 1953ndash70
Real R Pleguezuelos JM and Fahd S 1997 The distribution pat-terns of reptiles in the Riff region northern Morocco AfricanJournal of Ecology 35 312ndash325
Saeligtersdal M and Birks HJB 1993 Assessing the representative-ness of nature reserves using multivariate analysis vascularplants and breeding birds in deciduous forest western Norway
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 47
Biological Conservation 65 121ndash132Sibley CS and Monroe BE 1990 Distribution and Taxonomy of
Birds of the World Yale University Press New HavenSneath PHA and Sokal RR 1973 Numerical Taxonomy The
Principles and Practices of Numerical Classification FreemanSan Francisco
Sokal RR and Rohlf FJ 1981 Biometry The Principles andPractice of Statistics in Biological Research WH Freeman andCompany New York
Turner JRG Lennon JJ and Lawrenson JA 1988 British birdspecies distributions and the energy theory Nature 355 539ndash541
Turpie J K and Crowe TM 1994 Patterns of distribution diversityand endemism of larger African mammals South AfricanJournal of Zoology 29 19ndash32
Whitehead PJ Bowman DMJS and Tideman SC 1992Biogeographic patterns environmental correlates and conserva-tion of avifauna in the Northern Territory Australia Journal ofBiogeography 19 151ndash161
Wiens JA 1989 Spatial scaling in ecology Functional Ecology 3385ndash397
Williams PH De Klerk HM and Crowe TM 1999 Interpreting bio-geographical boundaries among Afrotropical birds spatial pat-terns in richness gradients and species replacement Journal ofBiogeography 26 459ndash474
Williams P Gibbons D Margules C Rebelo A Humphries C andPressey R 1996 A comparison of richness hotspots rarityhotspots and complementary areas for conserving diversity ofBritish birds Conservation Biology 10 155ndash174
World Resources Institute 1994 World Resources 1994ndash95 DataBase Diskette World Resources Institute Washington DC
Wright GH 1983 Species-energy theory an extension to species-area theory Oikos 41 496ndash506
Resumen
Se ha estudiado el aacuterea de distribucioacuten por paiacuteses de las 23 especies de caacutelaos africanos con el fin de identificar grupos dedistribuciones significativamente similares es decir corotipos y distribuciones que se reemplazan gradualmente Se hautilizado un meacutetodo probabiliacutestico de clasificacioacuten para distinguir entre ambos tipos de patrones biogeograacuteficos Una vezobtenidos los corotipos se analizoacute como se ven afectados por la influencia del clima el aacuterea la superficie cubierta por losdiferentes tipos de haacutebitat y las perturbaciones humanas en cada paiacutes Tambieacuten se comproboacute si las caracteriacutesticas propiasde cada especie entre las que se encuentran variables relacionadas con el haacutebitat la alimentacioacuten su biometriacutea y diferentesparaacutemetros metaboacutelicos influyen en los patrones biogeograacuteficos detectados Se han encontrado 3 corotipos que incluyen a20 especies El corotipo centro-occidental estaacute formado por 9 especies caracterizadas por la preferencia por haacutebitats debosque y por el bajo dimorfismo sexual y por 25 paiacuteses caracterizados por presentar un rango de temperaturas muy bajoEl corotipo oriental incluye a tres especies que se caracterizan por presentar haacutebitos alimenticios terrestres y un elevadodimorfismo sexual y a 7 paiacuteses caracterizados por mostrar gran parte de su superficie cubierta por sabanas y pastos Elcorotipo meridional agrupa la distribucioacuten de 8 especies que se caracterizan por su preferencia por bosques de montantildea ycosteros y zonas arboladas y a 18 paiacuteses con temperaturas bajas en el mes maacutes friacuteo y grandes superficies cubiertas pormatorrales y bosques de desierto Hay 3 especies que no se agrupan en corotipos y que siguen un reemplazamiento graduala lo largo del aacuterea de estudio
Chorotype 1African Pied Hornbill Tockus fasciatusPiping Hornbill Bycanistes fistulatorRed-billed Dwarf-hornbill Tockus camurusBlack-casqued Hornbill Ceratogymna atrataBrown-cheeked Hornbill Bycanistes cylindricusLong-tailed Hornbill Tropicranus albocristatusBlack Dwarf-hornbill Tockus hartlaubiGrey-cheeked Hornbill Bycanistes subcylindricusYellow-casqued Hornbill Ceratogymna elata
Chorotype 2Eastern Yellow-billed Hornbill Tockus flavirostrisVon der Deckenrsquos Hornbill Tockus deckeniHemprichrsquos Hornbill Tockus hemprichii
Appendix 1 Biotic elements of the three chorotypes obtained in Africa for the hornbills The corresponding geographic elements are shownin Figures 2 3 and 4 The species nomenclature follows Kemp (2001)
Chorotype 3Southern Ground-hornbill Bucorvus leadbeateriCrowned Hornbill Tockus alboterminatusPale-billed Hornbill Tockus pallidirostrisTrumpeter Hornbill Bycanistes bucinatorSilvery-cheeked Hornbill Bycanistes brevisBradfieldrsquos Hornbill Tockus bradfieldiSouthern Yellow-billed Hornbill Tockus leucomelasMonteirorsquos Hornbill Tockus monteiri
Ungrouped speciesAfrican Grey Hornbill Tockus nasutusRed Billed Hornbill Tockus erythrorhynchusNorthern Ground-hornbill Bucorvus abyssinicus
Received August 2001 accepted July 2002Editor WRJ Dean
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Muntildeoz Real Olivero Maacuterquez Guerrero Baacutercena and Vargas46
also singled out because of their movements T nasutus hasbeen reported to make regular movements near the Saharaand irregular irruptive movements in southern Africa (Kemp2001) while T erythrorhynchus appears to undertake regu-lar movements all over its distribution area In contrast theNorthern Ground-hornbill Bucorvus abyssinicus resides inthe sub-Saharan African savannas north of the equator(Kemp 1995)
These three species are the only ones occurring inMauritania Mali Niger and Burkina Faso which representthe northernmost distribution for the hornbills in Africa andT nasutus is the only hornbill that is also present in theArabian Peninsula It is note-worthy that if only the Africandistribution of the African Grey Hornbill is considered thenthese three species are grouped together constituting afourth chorotype and then all African hornbill species followa discrete distribution pattern in a strictly African contextHowever the presence of Tockus nasutus in south-east Asiaseems to have somehow disrupted the well established dis-crete biogeographical configuration of the African hornbills
Acknowledgements mdash We are grateful to Dr Alan Kemp Dr PilaiPoonswad Dr Helen de Klerk and Dr Juan Carranza for their criti-cal comments on an early version of this paper This study was fund-ed by the European Commision and Spanish Government project1FD97ndash1571ndashC02ndash01
References
Austin MP and Margules CR 1986 Assessing representativenessIn Usher MB (ed) Wildlife Conservation Evaluation pp 45ndash68Chapman and Hall London
Austin MP and Smith TM 1989 A new model for the continuumconcept Vegetatio 83 35ndash47
Baroni-Urbani C and Buser MW 1976 Similarity of binary dataSystematic Zoology 25 251ndash259
Baroni-Urbani C Rufo S and Vigna-Taglianti A 1978 Materialiper una biogeografia italiana fondata su alguni generi dicoleotteri cicindelidi carabidi e crisomelidi Estratto dellaMemorie della Societa Entomologica Italiana 56 35ndash92
Birks HJB 1987 Recent methodological development in quantita-tive descriptive biogeography Annales Zoology Fennici 24165ndash178
Brown JH 1995 Macroecology The University of Chicago PressChicago
Brown JH 1999 Macroecology progress and prospect Oikos 873ndash14
Cordellier S and Didiot B 1999 Lrsquoeacutetet du monde Eacuteditions LaDeacutecouverte et Syros Pariacutes
Crowe TM and Crowe AA 1982 Patterns in distribution diversityand endemism in Afrotropical birds Journal of Zoology London198 417ndash442
Crowe TM and Kemp AC 1988 African historical biogeography asreflected by Galliform and Hornbill evolution In Ouellet H (ed)Acta XIX Congressus Internationalis Ornithologici Vol 2 pp2510ndash2518 University of Ottawa Press Ottawa
Dasmann RF 1972 Towards a system for classifying naturalregions of the world and their representation by National Parksand Reserves Biological Conservation 4 247ndash255
Elmberg J Sjoumlberg K Nummi P and Poumlysauml H 1994 Patterns oflake acidity and waterfowl communities Hydrobiologia 279280201ndash206
FAO 1984 Agroclimatological data for Africa FAO Plant Productionand Protection Series No 22 Vols 1 and 2 Rome
Guillet A and Crowe TM 1985 Patterns of distribution speciesrichness endemism and guild composition of water-birds inAfrica African Journal of Ecology 23 89ndash120
Guillet A and Crowe TM 1986 A preliminary investigation of pat-terns of distribution and species richness of southern Africanwaterbirds South African Journal of Wildlife Research 1665ndash81
Hengeveld R 1990 Dynamic Biogeography Cambridge UniversityPress Cambridge
Hengeveld R and Hogeweg P 1979 Cluster analysis of the distri-bution patterns of Dutch carabid species (Col) In Orloci L RaoCR and Stiteler WM (eds) Multivariate Methods in EcologicalWork pp 65ndash86 International Co-operative Publishing HouseBurtonsville
Hildeacuten O 1965 Habitat selection in birds a review AnnalesZoology Fennici 2 53ndash75
Hosmer DW and Lemeshow S 1989 Applied Logistic RegressionJohn Wiley and Sons Inc New York
Kemp A 1995 The Hornbills Bucerotiformes Oxford UniversityPress
Kemp A 2001 Family Bucerotidae (Hornbills) In Del Hoyo J ElliotA and Sargatal J (eds) Handbook of the Birds of the World Vol6 Mousebirds to Hornbills pp 436ndash523 Lynx Editions Barcelona
Kikkawa J and Pearse K 1969 Geographical distribution of landbirds in Australia a numerical analysis Australian Journal ofZoology 17 821ndash840
Lance GN and Williams WT 1967 A general theory of classificato-ry sorting strategies I Hierarchical systems Computer Journa l9 373ndash380
Margules CR 1986 Conservation evaluation in practice In UsherMB (ed) Wildlife Conservation Evaluation pp 298ndash314 Chapmanand Hall London
Maacuterquez AL Real R Vargas JM and Salvo AE 1997 On identify-ing common distribution patterns and their causal factors a prob-abilistic method applied to pteridophytes in the Iberian PeninsulaJournal of Biogeography 24 613ndash631
Mayr E 1965 What is a fauna Zoologisches Jahrbuch derSystematik 92 473ndash486
McCoy ED Bell SS and Walters K 1986 Identifying bioticboundaries along environmental gradients Ecology 67 749ndash759
Olivero J Real R and Vargas JM 1998 Distribution of breedingwintering and resident waterbirds in Europe biotic regions andthe macroclimate Ornis Fennica 75 153ndash175
Rabinovich JE and Rapoport EH 1975 Geographical variation ofdiversity in passerine birds Journal of Biogeography 2141ndash157
Ramiacuterez JM and Vargas JM 1992 Contribucioacuten de la biogeografiacuteaa la gestioacuten del medio ambiente y a la conservacioacuten de lasespecies In Vargas JM Real R and Antuacutenez A (eds) Objetivos yMeacutetodos Biogeograacuteficos Aplicaciones en Herpetologiacutea 2 pp95ndash106 Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Vargas JM and Guerrero JC 1992a Anaacutelisis biogeograacuteficode clasificacioacuten de aacutereas y especies In Vargas JM Real R andAntuacutenez A (ed) Objetivos y Meacutetodos Biogeograacuteficos Aplicacionesen Herpetologiacutea Monografias Herpetoloacutegica 2 pp 73ndash84Asociacioacuten Herpetoloacutegica Espantildeola Valencia
Real R Guerrero JC and Ramiacuterez JM 1992b Identificacioacuten defronteras bioacuteticas significativas para los anfibios en la cuencahidrograacutefica del Sur de Espantildea Dontildeana Acta Vertebrata 1953ndash70
Real R Pleguezuelos JM and Fahd S 1997 The distribution pat-terns of reptiles in the Riff region northern Morocco AfricanJournal of Ecology 35 312ndash325
Saeligtersdal M and Birks HJB 1993 Assessing the representative-ness of nature reserves using multivariate analysis vascularplants and breeding birds in deciduous forest western Norway
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 47
Biological Conservation 65 121ndash132Sibley CS and Monroe BE 1990 Distribution and Taxonomy of
Birds of the World Yale University Press New HavenSneath PHA and Sokal RR 1973 Numerical Taxonomy The
Principles and Practices of Numerical Classification FreemanSan Francisco
Sokal RR and Rohlf FJ 1981 Biometry The Principles andPractice of Statistics in Biological Research WH Freeman andCompany New York
Turner JRG Lennon JJ and Lawrenson JA 1988 British birdspecies distributions and the energy theory Nature 355 539ndash541
Turpie J K and Crowe TM 1994 Patterns of distribution diversityand endemism of larger African mammals South AfricanJournal of Zoology 29 19ndash32
Whitehead PJ Bowman DMJS and Tideman SC 1992Biogeographic patterns environmental correlates and conserva-tion of avifauna in the Northern Territory Australia Journal ofBiogeography 19 151ndash161
Wiens JA 1989 Spatial scaling in ecology Functional Ecology 3385ndash397
Williams PH De Klerk HM and Crowe TM 1999 Interpreting bio-geographical boundaries among Afrotropical birds spatial pat-terns in richness gradients and species replacement Journal ofBiogeography 26 459ndash474
Williams P Gibbons D Margules C Rebelo A Humphries C andPressey R 1996 A comparison of richness hotspots rarityhotspots and complementary areas for conserving diversity ofBritish birds Conservation Biology 10 155ndash174
World Resources Institute 1994 World Resources 1994ndash95 DataBase Diskette World Resources Institute Washington DC
Wright GH 1983 Species-energy theory an extension to species-area theory Oikos 41 496ndash506
Resumen
Se ha estudiado el aacuterea de distribucioacuten por paiacuteses de las 23 especies de caacutelaos africanos con el fin de identificar grupos dedistribuciones significativamente similares es decir corotipos y distribuciones que se reemplazan gradualmente Se hautilizado un meacutetodo probabiliacutestico de clasificacioacuten para distinguir entre ambos tipos de patrones biogeograacuteficos Una vezobtenidos los corotipos se analizoacute como se ven afectados por la influencia del clima el aacuterea la superficie cubierta por losdiferentes tipos de haacutebitat y las perturbaciones humanas en cada paiacutes Tambieacuten se comproboacute si las caracteriacutesticas propiasde cada especie entre las que se encuentran variables relacionadas con el haacutebitat la alimentacioacuten su biometriacutea y diferentesparaacutemetros metaboacutelicos influyen en los patrones biogeograacuteficos detectados Se han encontrado 3 corotipos que incluyen a20 especies El corotipo centro-occidental estaacute formado por 9 especies caracterizadas por la preferencia por haacutebitats debosque y por el bajo dimorfismo sexual y por 25 paiacuteses caracterizados por presentar un rango de temperaturas muy bajoEl corotipo oriental incluye a tres especies que se caracterizan por presentar haacutebitos alimenticios terrestres y un elevadodimorfismo sexual y a 7 paiacuteses caracterizados por mostrar gran parte de su superficie cubierta por sabanas y pastos Elcorotipo meridional agrupa la distribucioacuten de 8 especies que se caracterizan por su preferencia por bosques de montantildea ycosteros y zonas arboladas y a 18 paiacuteses con temperaturas bajas en el mes maacutes friacuteo y grandes superficies cubiertas pormatorrales y bosques de desierto Hay 3 especies que no se agrupan en corotipos y que siguen un reemplazamiento graduala lo largo del aacuterea de estudio
Chorotype 1African Pied Hornbill Tockus fasciatusPiping Hornbill Bycanistes fistulatorRed-billed Dwarf-hornbill Tockus camurusBlack-casqued Hornbill Ceratogymna atrataBrown-cheeked Hornbill Bycanistes cylindricusLong-tailed Hornbill Tropicranus albocristatusBlack Dwarf-hornbill Tockus hartlaubiGrey-cheeked Hornbill Bycanistes subcylindricusYellow-casqued Hornbill Ceratogymna elata
Chorotype 2Eastern Yellow-billed Hornbill Tockus flavirostrisVon der Deckenrsquos Hornbill Tockus deckeniHemprichrsquos Hornbill Tockus hemprichii
Appendix 1 Biotic elements of the three chorotypes obtained in Africa for the hornbills The corresponding geographic elements are shownin Figures 2 3 and 4 The species nomenclature follows Kemp (2001)
Chorotype 3Southern Ground-hornbill Bucorvus leadbeateriCrowned Hornbill Tockus alboterminatusPale-billed Hornbill Tockus pallidirostrisTrumpeter Hornbill Bycanistes bucinatorSilvery-cheeked Hornbill Bycanistes brevisBradfieldrsquos Hornbill Tockus bradfieldiSouthern Yellow-billed Hornbill Tockus leucomelasMonteirorsquos Hornbill Tockus monteiri
Ungrouped speciesAfrican Grey Hornbill Tockus nasutusRed Billed Hornbill Tockus erythrorhynchusNorthern Ground-hornbill Bucorvus abyssinicus
Received August 2001 accepted July 2002Editor WRJ Dean
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Ostrich 2003 74 39ndash47 47
Biological Conservation 65 121ndash132Sibley CS and Monroe BE 1990 Distribution and Taxonomy of
Birds of the World Yale University Press New HavenSneath PHA and Sokal RR 1973 Numerical Taxonomy The
Principles and Practices of Numerical Classification FreemanSan Francisco
Sokal RR and Rohlf FJ 1981 Biometry The Principles andPractice of Statistics in Biological Research WH Freeman andCompany New York
Turner JRG Lennon JJ and Lawrenson JA 1988 British birdspecies distributions and the energy theory Nature 355 539ndash541
Turpie J K and Crowe TM 1994 Patterns of distribution diversityand endemism of larger African mammals South AfricanJournal of Zoology 29 19ndash32
Whitehead PJ Bowman DMJS and Tideman SC 1992Biogeographic patterns environmental correlates and conserva-tion of avifauna in the Northern Territory Australia Journal ofBiogeography 19 151ndash161
Wiens JA 1989 Spatial scaling in ecology Functional Ecology 3385ndash397
Williams PH De Klerk HM and Crowe TM 1999 Interpreting bio-geographical boundaries among Afrotropical birds spatial pat-terns in richness gradients and species replacement Journal ofBiogeography 26 459ndash474
Williams P Gibbons D Margules C Rebelo A Humphries C andPressey R 1996 A comparison of richness hotspots rarityhotspots and complementary areas for conserving diversity ofBritish birds Conservation Biology 10 155ndash174
World Resources Institute 1994 World Resources 1994ndash95 DataBase Diskette World Resources Institute Washington DC
Wright GH 1983 Species-energy theory an extension to species-area theory Oikos 41 496ndash506
Resumen
Se ha estudiado el aacuterea de distribucioacuten por paiacuteses de las 23 especies de caacutelaos africanos con el fin de identificar grupos dedistribuciones significativamente similares es decir corotipos y distribuciones que se reemplazan gradualmente Se hautilizado un meacutetodo probabiliacutestico de clasificacioacuten para distinguir entre ambos tipos de patrones biogeograacuteficos Una vezobtenidos los corotipos se analizoacute como se ven afectados por la influencia del clima el aacuterea la superficie cubierta por losdiferentes tipos de haacutebitat y las perturbaciones humanas en cada paiacutes Tambieacuten se comproboacute si las caracteriacutesticas propiasde cada especie entre las que se encuentran variables relacionadas con el haacutebitat la alimentacioacuten su biometriacutea y diferentesparaacutemetros metaboacutelicos influyen en los patrones biogeograacuteficos detectados Se han encontrado 3 corotipos que incluyen a20 especies El corotipo centro-occidental estaacute formado por 9 especies caracterizadas por la preferencia por haacutebitats debosque y por el bajo dimorfismo sexual y por 25 paiacuteses caracterizados por presentar un rango de temperaturas muy bajoEl corotipo oriental incluye a tres especies que se caracterizan por presentar haacutebitos alimenticios terrestres y un elevadodimorfismo sexual y a 7 paiacuteses caracterizados por mostrar gran parte de su superficie cubierta por sabanas y pastos Elcorotipo meridional agrupa la distribucioacuten de 8 especies que se caracterizan por su preferencia por bosques de montantildea ycosteros y zonas arboladas y a 18 paiacuteses con temperaturas bajas en el mes maacutes friacuteo y grandes superficies cubiertas pormatorrales y bosques de desierto Hay 3 especies que no se agrupan en corotipos y que siguen un reemplazamiento graduala lo largo del aacuterea de estudio
Chorotype 1African Pied Hornbill Tockus fasciatusPiping Hornbill Bycanistes fistulatorRed-billed Dwarf-hornbill Tockus camurusBlack-casqued Hornbill Ceratogymna atrataBrown-cheeked Hornbill Bycanistes cylindricusLong-tailed Hornbill Tropicranus albocristatusBlack Dwarf-hornbill Tockus hartlaubiGrey-cheeked Hornbill Bycanistes subcylindricusYellow-casqued Hornbill Ceratogymna elata
Chorotype 2Eastern Yellow-billed Hornbill Tockus flavirostrisVon der Deckenrsquos Hornbill Tockus deckeniHemprichrsquos Hornbill Tockus hemprichii
Appendix 1 Biotic elements of the three chorotypes obtained in Africa for the hornbills The corresponding geographic elements are shownin Figures 2 3 and 4 The species nomenclature follows Kemp (2001)
Chorotype 3Southern Ground-hornbill Bucorvus leadbeateriCrowned Hornbill Tockus alboterminatusPale-billed Hornbill Tockus pallidirostrisTrumpeter Hornbill Bycanistes bucinatorSilvery-cheeked Hornbill Bycanistes brevisBradfieldrsquos Hornbill Tockus bradfieldiSouthern Yellow-billed Hornbill Tockus leucomelasMonteirorsquos Hornbill Tockus monteiri
Ungrouped speciesAfrican Grey Hornbill Tockus nasutusRed Billed Hornbill Tockus erythrorhynchusNorthern Ground-hornbill Bucorvus abyssinicus
Received August 2001 accepted July 2002Editor WRJ Dean
Dow
nloa
ded
by [
UM
A U
nive
rsity
of
Mal
aga]
at 0
115
06
Sept
embe
r 20
11
Related Documents
