Coincidence and mismatch of biodiversity hotspots

Coincidence and mismatch of biodiversity hotspots

Biological Conservation 93 (2000) 163±175 Coincidence and mismatch of biodiversity hotspots: a global survey for the ...

1MB Sizes 3 Downloads 6 Views

Biological Conservation 93 (2000) 163±175

Coincidence and mismatch of biodiversity hotspots: a global survey for the order, primates A.H. Harcourt Department of Anthropology, University of California, One Shields Ave., Davis, CA 95616, USA Received 16 June 1999; accepted 14 September 1999

Abstract A global survey of a well-studied order of tropical mammals, primates, is used to explore the use of diversity hotspots in conservation. The results at this shallow taxonomic level match those for most cross-phylum analyses. Overlap of hotspots for species, genera, trait-complexes, families, and threatened species varies with the continent, and the comparison. Overlap is best in Africa and Madagascar, and poorest in Asia, but reasons for the di€erences need exploring. A complete mismatch of taxonomic and threatened species hotspots in South America, resulting from the mismatch of the hotspots of diversity and human destruction, suggests that conservation biologist's hotspot approach could bene®t from adding hotspots of human threat to the analysis of diversity hotspots. Conservationists' use of hotspot analysis seems to have been largely empirical. If analysis of single orders can contribute to conservation, application of biogeographic theory to our knowledge of the distribution of the order, primates, is the next step. # 2000 Elsevier Science Ltd. All rights reserved. Keywords: Biodiversity; Biogeography; Hotspot; Primates; Threatened species

1. Introduction `Hotspots', regional concentrations of species, have been of interest to biogeographers since the early 1800s (Browne, 1983). They are also of interest to conservationists, because of their potential to provide easy identi®cation of sites for preservation of biodiversity (Terborgh and Winter, 1983; Myers, 1988; Prendergast et al., 1993; Pressey et al., 1993; Scott et al., 1993; Dobson et al., 1997; Mittermeier et al., 1998; Reid, 1998). However, while some areas of high diversity overlap, often they do not. Thus in the USA, the southeast is an overall hotspot, but endangered plants are largely in the west (Dobson et al., 1997; Ricketts et al., 1999). In Transvaal, South Africa, little overlap exists of hotspots or coldspots among plant and animal phyla, or even among rare representatives of those phyla (van Jaarsveld et al., 1998). In Britain, the maximum overlap between any two taxa is only 34% (butter¯ies and dragon¯ies), and two charismatic taxa, birds and butter¯y, overlap by only 12% (Prendergast et al., 1993). Even E-mail address: [email protected] (A.H. Harcourt).

among the relatively similar Ugandan forests, only three of 10 comparisons across ®ve phyla were signi®cantly congruent once area of forest was accounted for (Howard et al., 1998). Thus from identifying sites that might contain hotspots for several taxa, the hotspot approach has moved more strongly to identifying complementary sites, namely the range of sites that represent as much as possible of the range of taxa (Pressey et al., 1993; Williams et al., 1997). However, a potential complication for generalisation across continents is the ®nding that the extent of overlap (or lack of it) that exists between particular taxa varies with continent (Pearson and Carroll, 1998). As the examples given above indicate, coincidence and mismatch of hotspots tends to be sought among classes or phyla. The greater the number of phyla with overlapping hotspots, the less area needs conserving to protect biodiversity, and thus the more likely biodiversity is to be protected. In addition, the practical hope was that hotspots for easily seen and identi®ed charismatic taxa would match those for less obvious taxa. Far less frequently are the regional comparisons done at shallower taxonomic levels, except at the local scale; the reason is that we do not expect concordance at shallow

0006-3207/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S0006-3207(99)00145-7


A.H. Harcourt / Biological Conservation 93 (2000) 163±175

taxonomic levels (Reid, 1998). Hacker et al. (1998) recently provided one of the few applications of the hotspot approach within an order, the primates. They showed that hotspots for threatened primate subspecies in Africa and Madagascar together were hotspots for all primate subspecies, but that little overlap existed between hotspots of endemics (rare subspecies) and either threatened subspecies or all subspecies, because endemic hotspots were scattered. By comparison to many other taxa, the order primates consists of mostly large, easily observable species, even the nocturnal ones. They are thus well-studied (Rowe, 1996). Consequently, we probably have as good information on the global distribution of primate taxa as we do for any other tropical order. Such information is particularly important, because tropical countries have generally been less well surveyed and currently have less funding for surveys than do temperate countries (Myers, 1984; Harcourt, 1995; Howard et al., 1998). Primates could thus be good indicator species for countries with less opportunity for biological surveys than others. Understanding of the distribution of their diversity could therefore be especially useful. In order to further explore the generality of the hotspot approach, I here provide the ®rst, detailed global assessment of coincidence and mismatch of hotspots for the order, primates, at several taxonomic levels of analysis, treating each continent separately. Hotspots for di€erent taxonomic levels are compared, as are hotspots for di€erent trait-complexes, including threatened species. The main questions concern how well the generic hotspots represent the speci®c ones; how well either of these represent the various trait complexes; and ®nally, how well the threatened species' hotspots overlap the various taxonomic hotspots. In other words, can we use deeper taxa as indicators? Does the taxon's hotspots of overall diversity represent the variety within the taxon?; and will conservation of threatened taxa conserve the full variety of the taxon? 2. Methods 2.1. Sources of data The basis of this analysis are the distribution maps [`coverages' in geographic information systems (GIS) terminology] of individual species [N=about 60 species per continent, except Madagascar with 30 (see Figs. 1±4)]. These were digitised from maps in Wolfheim (1983), with subsequent correction and addition from, in particular, Niemitz (1984), Hershkovitz (1987a; 1987b; 1990), Lernould (1988), Nash, Bearder and Olson (1989), Harcourt and Thornhill (1990), Corbet and Hill (1992), Groves (1993), Rylands, Coimbra-Filho and Mittermeier (1993), Ford (1994), Mittermeier et al.

(1994), Oates, Davies and Delson (1994), and Kinzey (1997). The `Regions' feature class of ARC/INFO (ESRI Inc., 1998a) was then used to draw `polygons'; that delineated regions with di€ering numbers of species. The result is areas of concentration of those di€ering numbers of species. Taxonomic nomenclature and listing is mostly from Corbet and Hill (1991) and Groves (1993). Where the sources di€ered, the one with the fewer species was chosen, although Papio remains as ®ve species. Callicebus is treated as three species (as in Wolfheim 1983), because the two main authorities di€er (Hershkovitz, 1990; Groves, 1993). Threatened taxa are those listed in the 1996 IUCN Red List of Threatened Animals (IUCN, 1996) as Vulnerable, Endangered, and Critically Endangered. While the categorisation of status in the Red List is challenged for some taxa (Mrosovsky, 1997), and con¯icts with published, substantiated analyses for others (Harcourt, 1996), nevertheless, it is the most complete source available, and heavily used in national conservation policy. Not all taxa in the IUCN List were included in the analysis here. Missing species are those that do not appear in one or other of the taxonomic sources. The missing species are largely recent splits re¯ecting minor morphological variations at single geographic sites, some of which are taxonomically contested. The exclusion of the IUCN species in no instance a€ects the IUCN classi®cation of their sister species, because the geographic ranges of the excluded species are very small by comparison to their sister species' ranges. Brachyteles is the one exception: the split species each have about the same size of geographic range. However, the original single species is as threatened as either of IUCN's two species. The IUCN species not included in this analysis are, in IUCN order of listing, Aotus brumbacki, A. lemurinus, Brachyteles hypoxanthus, Callicebus dubius, Cebus kaapori, C. xanthosternos, Cercopithecus preussi, C. sclateri, C. solatus, Hapalemur aureus, Macaca brunnescens, M. pagensis, Microcebus myoxinus, Saimiri oerstedii (possibly a human introduction), S. vanzolinii, Trachypithecus delacouri, and T. poliocephalus. The maps are displayed in ArcView's (3.1) Geographical projection (ESRI Inc., 1998b). 2.2. De®nitions of hotspots No one de®nition or mode of measurement of a hotspot seems to be preferred. Here, I use overlap of the distribution maps of individual taxa (not the common counts of taxa in quadrats). Where the sample size is large enough, four levels of concentration of taxa are shown. The four levels are as near as possible in quartiles of the number of taxa with overlapping distributions, starting from the lowest quartile, but the exact

A.H. Harcourt / Biological Conservation 93 (2000) 163±175

proportion of taxa in the top quartile varies depending on the number of taxa in the sample. A `hotspot' is de®ned here as the top quartile (or the top half for threatened species).


2.3. Analysis Where hotspot overlap is concerned, two main measures are given. The proportion of a hotspot that is

Fig. 1. Map of concentrations of primate taxa separated into quartiles of numbers of species. Africa (omitting northern Africa and the Barbary macaque, Macaca sylvanus).


A.H. Harcourt / Biological Conservation 93 (2000) 163±175

overlapped by another hotspot (e.g., the proportion of the species' hotspot that is overlapped by the generic hotspot), and the proportion of taxa that are overlapped by a hotspot (e.g., the proportion of species

overlapped by the generic hotspot). The former measure is, in e€ect, the probability of a point site in a hotspot hitting the other hotspot. By de®nition, 25% of the taxa of the overlapped taxonomic category will be at that

Fig. 2. Map of concentrations of primate taxa separated into quartiles of numbers of species. Asia (omitting Japan and the Japanese macaque, Macaca fuscata). The omissions of the two Macaca are merely for reasons of space on the page. The species were included in analysis, and neither omission in presentation a€ects the resulting hotspots.

A.H. Harcourt / Biological Conservation 93 (2000) 163±175

point site. In contrast, often more than 25% of the species of the overlapped category will be overlain, because the identity of the species di€er from point to point within a hotspot, even though there will be the same proportion of the total across the hotspot. The usual use of hotspots is as guidance for conservation e€ort.


Almost all the hotspots are orders of magnitude larger than the average current protected area. For instance, Africa's mean current reserve size is 2000 km2 (Miller et al., 1995), whereas its species' and threatened species' hotspots are each over half a million square kilometres in total. Thus, a point site is going to be closer in size to

Fig. 3. Map of concentrations of primate taxa separated into quartiles of numbers of species. Madagascar.


A.H. Harcourt / Biological Conservation 93 (2000) 163±175

the average reserve than is the size of a whole hotspot. The number of species within a reserve is thus going to be somewhere between the number at a point site, and the number overlapped by a whole hotspot, and probably far closer to the former.

3. Results Throughout, I treat the continents in alphabetical order. Hotspots for primate taxa in the four continents are obvious, whatever the level of analysis (Figs. 1±5).

Fig. 4. Map of concentrations of primate taxa separated into quartiles of numbers of species. South America.

A.H. Harcourt / Biological Conservation 93 (2000) 163±175

As many others have shown, primate hotspots cluster at the equator, except in Madagascar. They also cluster in forests, the habitat of most primate species (Wallace, 1876, Ch. 17; Terborgh and Schaik, 1987), and in areas of high rainfall, except in Asia (Reed and Fleagle, 1995).


Africa's hotspots contain the highest number of species, despite Asia and South America having very similar numbers of species in total. Nevertheless, conserving just Africa's hotspots will not conserve a representative sample of all primates. As Wallace (1876, Vol.2, p.179)

Fig. 5. Maps of concentrations of threatened primate taxa, in two categories of numbers of species. (a) Africa; (b) Asia (omitting Japan and the Japanese macaque, Macaca fuscata); (c) Madagascar; (d) South America.


A.H. Harcourt / Biological Conservation 93 (2000) 163±175

that more variety is probably contained in a generic hotspot than in a speci®c one, given that two genera are on average more di€erent than are any two species within those genera. However, the relative diversity within di€erent taxonomic levels varies according to the identity of the taxa compared (Crozier, 1997). The relative size and number of the generic and speci®c hotspots varies across continents, but overlap of the speci®c hotspots by the generic hotspots is 75% or more for all continents, except Asia (Figs. 1±4; Table 2). Overlap of the speci®c by the generic hotspots is highest in Africa, where the generic and speci®c hotspots' overlap with the subspeci®c hotspots is also high [compare Fig. 1a,b with Hacker et al.'s (1998) Fig 2a]. In Asia, a concentration of e€ort on generic hotspots would lead to conservation in only Borneo, so missing the mainland speci®c hotspots (Fig. 2 a,b). Generic hotspots cover about 50% of the number of species for all continents, except again Asia, whereas the speci®c hotspots cover 75% or more of genera on all continents (Table 3). The generic hotspot covers over 50% of the area of the threatened hotspots in Africa and Madagascar, compared to none of them in Asia or South America (Table 2). However, for none of the continents does the generic hotspot cover substantially more threatened species than does the speci®c hotspot, and indeed in Asia it covers a third of the number of threatened species (Table 3). For primates, therefore, genera are not useful indicator taxa in a hotspot application for conservation.

wrote, `The most striking fact presented by this order . . . is the strict limitation of well-marked families to de®nite areas'. Baboons, for example, occur in only Africa, gibbons in only Asia, lemurs in only Madagascar, and marmosets in only South America. Because the taxa di€er across the continents, so too do the traits that they represent (Terborgh and Schaik, 1987; Maurer et al., 1992; Kappeler and Heymann, 1996; Williams and Humphries, 1996). In all the continents, a relatively small proportion of the total area covered by primates (<10%) contains a relatively large proportion of the total number of taxa in that continent (Table 1). Thus the speci®c hotspots contain 40±62% of species (Table 1), and the generic hotspots, 67±93% of genera. However, the Lemuridae in Madagascar illustrate the di€erence that de®nitions can make to the size of a hotspots, and therefore in inferences from hotspots' analysis. The Lemuridae's top quartile hotspot could reasonably be de®ned as either the 5-species zone (as here), or as the 4±5 species zone. The 5-species hotspot covers 275 km2; the 4±5 species zone covers 28 350 km2, a 100-fold di€erence (Fig. 3g). 3.1. Deeper taxa as indicator taxa Can deeper taxa (genera rather than species, families rather than genera) be used as indicator taxa? In this paper I ask whether generic hotspots include most species. Deeper taxa have sometimes been found to be adequate indicators (Williams and Gaston, 1994), and sometimes not so found (van Jaarsveld et al., 1998). However, the studies di€ered not only in the region studied, but also in their de®nition of what a hotspot was (as is often the case in the hotspot literature). Another reason to concentrate on generic hotspots is

3.2. Hotspots for trait-complexes A number of authors have raised the issue of the importance of diversity of form, as well as of number of taxa (Janzen, 1988; Jablonski, 1995; Dingle et al., 1997;

Table 1 Percent total geographic area and percent species of primates per continent covered by ROW hotspots Africa



South America

% Area

% Species

% Area

% Species

% Area

% Species

% Area

% Species

Species Genera Nocturnal Diurnal For. Cerc.a Colobinae Apes Cheirogaleidae Indriidae Lemuridae Callitrichidae Cebinae Aotinae, etcb

1.6 3.2 0.4 3.4 1.2 2.5 ± ± ± ± ± ± ±

62 66 55 71 67 55 ± ± ± ± ± ± ±

1.7 1.2 8.2 0.5 ± ± 0.3 ± ± ± ± ± ±

40 22 50 31 ± ± 17 ± ± ± ± ± ±

7.5 5.6 0.0 4.3 ± ± ± 5.2 5.9 0.1 ± ± ±

53 50 33 47 ± ± ± 53 43 37 ± ± ±

4.9 10.1 ± ± ± ± ± ± ± ± 0.1 1.6 6.6

41 47 ± ± ± ± ± ± ± ± 35 24 47










a b

For. Cerc., Forest Cercopithecine. Aotinae etc., Aotinae, Pithecinae, Atelinae.

A.H. Harcourt / Biological Conservation 93 (2000) 163±175

Hacker et al., 1998; Jernvall and Wright, 1998; see also Shepherd, 1998). Thus Williams and Humphries (1996) showed for Bombus bees, that while the hotspot for number of species was in South America, the hotspot for number of individual traits was in central and eastern Eurasia, and the hotspot for number of combinations of characters was in eastern Eurasia. For African primates, Hacker et al. (1998) showed a good match of taxonomic and character richness across Africa. Madagascar, however, has only one major taxon, the Lemuroidea, and their character richness index was derived from taxonomic variety, so causing a mismatch between subspecies richness and character richness (Hacker et al., 1998). The trait-complex examined in this paper, for purposes of illustration, is activity period. This traitcomplex correlates strongly with body mass and sociality: nocturnal taxa are usually smaller and more solitary than are diurnal species (Clutton-Brock and Harvey, 1977). The trait-complex is also a taxonomic grouping in Africa and Asia, because the the nocturnal taxa tend to be prosimians. In Madagascar, all the primates are prosimians, and there the two activity periods (and their associated traits) are not distributed randomly among the taxa. South America has only one nocturnal taxon, Aotus, and is thus excluded. The overlap of speci®c and generic hotspots with trait-complex hotspots, and the overlap among trait-


complex hotspots varies with continent (Figs. 1±3; Tables 2 and 3). The match was most consistent in Africa, and relatively high, as was the match between subspeci®c and character richness in Hacker et al.'s (1998) analysis. However, the western hotspots of nocturnal and diurnal primates miss one another (Fig. 1). In Asia, large contrasts in the size of the hotspots, and the con®nement of the generic hotspots to Borneo, produce imbalance in overlap of both area and number of taxa. In Madagascar, both speci®c and generic hotspots overlap well the diurnal hotspots, but there is complete mismatch of the nocturnal and diurnal hotspots themselves. However, the species and generic hotspots in Madagascar overlay a large (>50%) proportion of both the nocturnal and diurnal species (Table 3). 3.3. Hotspots for families The more restricted the taxonomic grouping, the less likely is it that hotspots will coincide, for both historical and environmental biogeographic reasons (Williams and Humphries, 1996; Reid, 1998). Nevertheless, within one mammalian order that is largely con®ned to one habitat, tropical forest, considerable overlap of family hotspots by speci®c or generic hotspots might be expected. Families can also be considered as trait-complexes.

Table 2 Summary of overlap of area of hotspotsa Africa


Hotspots Species Genera Threatened

Species ± 93 51




For. Cerc.b 23.5 19.5 2

Colobine 5 3 14

Genera 45.5 ± 45.5

Nocturnal 57 73 28

Diurnal 45 46 23



± 31 61

45 ± 0

Diurnal±Mon- Apes keys 99 0 20 0 76 0


Species Genera Threatened

Nocturnal± Prosim 19.5 11.5 43



Hotspots Species Genera Threatened

Species ± 75 100

Genera 100 ± 100

Nocturnal 0 15 100

Diurnal 99 87.5 100

Cheirogaleidae 0 0 2

Indriidae 94 95 99

South America


Hotspots Species Genera Threatened

Species ± 100 0

Genera 48 ± 0

Callitrichidae 100 100 0

Cebinae 0 11 0

Aotinae, etc.c 52 80 0

Threatened 0 0 ±


Threatened 40 73 ±

20 0 ±

Lemuridae 100 0 100

Threatened 77 58 ±

Entries, percent of COLUMN hotspot overlapped by ROW hotspot, or percent chance of point protected area in row hotspot overlapping column hotspot. b For. Cerc., Forest Cercopithecine. c Aotinae etc., Aotinae, Pithecinae, Atelinae.


A.H. Harcourt / Biological Conservation 93 (2000) 163±175

Overlap of the speci®c and generic hotspots with the familial hotspots is variable within as well as between continents. While no one continent stands out as obviously more or less consistent than any other, especially with regard to area of of overlap, a consistently large proportion of Madagascar's species from separate families are overlapped by both the speci®c and generic hotspots (Figs. 1±4; Tables 2 and 3). In addition, the familial hotspots themselves rarely show substantial overlap (Figs. 1±4). In Africa, the far, western colobine hotspot [foregut fermentation with lysozyme as the trait-complex (Messier and Stewart, 1997)] is an obvious outlier, but nevertheless the speci®c and generic hotspots overlap the ranges of over 50% of colobine species (Fig. 1; Tables 2 and 3). In Asia, the scattered speci®c hotspots overlap a greater number of familial hotspots regions than do the more con®ned generic hotspots (Fig. 2; Table 2). Although the ape hotspot is an obvious outlier, the speci®c hotspot overlaps the ranges of 50% of the ape species (Table 3). Perhaps the most noticeable phenomenon in Madagascar is the lack of any overlap by the speci®c or generic hotspots of the

cheirogaleid hotspot (Fig. 3; Table 2), yet the 100% overlap with number of cheirogaleid species (Table 3). In South America, the familial hotspots are o€set enough, and the sizes of the speci®c ranges di€erent enough, to produce confusing di€erences between hotspot overlap, and proportion of species overlapped by the hotspots (Table 2 vs 3). 3.4. Threatened species hotspots Threatened taxa, and ecosystems are a non-random selection of all taxa and ecosystems (Jernvall and Wright, 1998). Overlap of either speci®c or generic hotspots with the hotspots for threatened species is therefore not necessarily expected, and in some cases not found (Prendergast et al., 1993). At the same time, a concentration of species with small geographic ranges could form a speci®c hotspot, which is then necessarily a threatened species hotspot because species with small geographic ranges are threatened (Brown, 1995, p. 215). Also, primates are concentrated in tropical forest, and tropical forest is under threat (Hannah et al., 1994;

Table 3 Summary of taxonomic overlap of hotspotsa,b Africa


Hotspots Species Genera Threatened

Species n=58 62 65.5 59



Hotspots Species Genera Threatened

Species n=58 40 22 36



Hotspots Species Genera Threatened

Species n=30 53 50 70

South America



Species n=66 41 47 18

Species Genera Threatened a

Genera 21 81 86 86

Nocturnal 15 80 80 60

Diurnal 42 62 64 62

For. Cerc.c 17 71 76.5 65

Colobine 7 57 57 71

Genera 12 75 67 58

Nocturnal±Prosim 7 43 29 29

Diurnal±Monkey 39 36 20 33

Apes 12 50 25 50

Threatened 31 29 10 ±

Genera 14 93 93 93

Nocturnal 17 76 76 88

Diurnal 13 61 54 77

Cheirogaleidae 7 100 100 100

Indriidae 14 64 64 79

Genera 16 75 81 37

Callitrichidae 29 31 34 28

Cebinae 5 60 80 20

Aotinae, etc.d 32 47 53 9

Threatened 20 20 20 ±

Threatened 10 50 50 ±

Lemuridae 9 56 44 78

Threatened 17 69 53 ±

Entries, percent of COLUMN taxa overlapped by ROW hotspot. Note for all continents except Madagascar, if overlap was less than 1000 km2 for any species, the species was omitted, because most species' ranges were far larger, and thus preserving that 1000 km2 would make little di€erence to persistence. In Madagascar, many species had extremely small ranges. c For. Cerc., Forest Cercopithecine. d Aotinae etc., Aotinae, Pithecinae, Atelinae. b

A.H. Harcourt / Biological Conservation 93 (2000) 163±175

Olson and Dinerstein, 1998). Thus, for African and Malagasay primates combined, Hacker et al.'s (1998) results indicate good overlap between threatened subspeci®c hotspots and total subspeci®c hotspots. Given the importance that we need to attach to threatened taxa, overlap of hotspots of overall diversity with threatened hotspots needs to be good if it is to be useful. In this global analysis, the overlap in either direction between speci®c or generic hotspots and threatened species hotspots is better in Africa, and especially Madagascar, than it is in Asia or South America (Figs. 1±4 vs 5; Tables 1±3). In Africa, about 50% overlap of hotspot areas occurred (Figs. 1 and 5a; Table 2), with a slightly higher percentage for proportion of species and genera overlain by the threatened hotspots (Table 3). Hacker et al. (1998) found substantial overlap in Africa between the threatened and total subspeci®c hotspots, although their analysis of subspecies produced an east African threatened hotspot, whereas this analysis has only central and west African threatened species hotspots (Fig. 5a). In Asia, the speci®c, and especially generic hotspots miss the northern hotspot of threatened mainland species (Figs. 2 and 5b), but despite the relatively high overlap of speci®c by threatened hotspots (Figs. 2 and 5b; Table 2), only the proportion of genera overlapped by the threatened hotspot exceeds 50% (Table 3). In Madagascar, as said, overlap in both directions of both area and number of taxa is high (Fig. 3; Tables 2 and 3). In South America, by contrast, a complete mismatch of hotspots occurs: the taxonomic hotspots (speci®c to familial) are in the west, but the threatened hotspots are in the south-eastern forests (Figs. 4 and 5d; Table 2). Correspondingly, the proportion of threatened species overlain by species or generic hotspots, or the proportion of species or genera overlain by the threatened hotspots is very low (Table 3). With respect to overlap of subtaxonomic hotspots and species by the threatened hotspots, the overlap of area of subtaxonomic hotspots, and especially of the overlap with number of species of those subtaxa, is also consistently better in Africa and Madagascar than in Asia or South America (Tables 2 and 3). 4. Discussion The nature of overlap of hotspots varies considerably among the continents. My speci®c to familial level analysis closely matches Hacker et al.'s (1998) subspeci®c analysis in indicating substantial overlap of the various hotspots in Africa and Madagascar. However, these continents are not models for the global distribution of primates. Asia especially, but also South America, have disjunct distributions of primates. Thus a hotspot approach to conservation of primates will in general work better in Africa and Madagascar than in Asia or


South America. That is, use of hotspots to position protected areas will overlap more trait-complexes, more taxa, and more threatened species in Africa and Madagascar than in Asia or South America. The most obvious geographic contrast between Asia and the other three continents is that Asia is a heavily divided landmass, which might correlate with its heavily divided distribution of primates (e.g., Brandon-Jones, 1996). South America shows the severest mismatch of taxonomic hotspots with hotspots for threatened species, and vice versa. Manne et al. (1999) found for South America the same mismatch between diversity hotspot and threatened species hotspot for passerine birds, and for the same reason: human disturbance is intense in the south-eastern, Brazilian, Atlantic coast forest, but not yet in the west (Hannah et al., 1994; Olson and Dinerstein, 1998). Conservation needs shortcuts because time and money are so lacking (Mittermeier et al., 1998). But precisely because of the constraints, shortcuts must be backed by strongly substantiated analysis (Leader-Williams and Albon, 1988). The hotspot analysis reported here for one order of mammals, primates, epitomises most, maybe all, hotspot analyses. De®nitions of hotspots massively a€ect their size and site, and hence degree of overlap. While some considerable overlap of hotspots across di€erent taxonomic (or other) categories exists, there is appreciable mismatch. And much of the mismatch is unexplained, including obvious differences between continents in degree and nature of overlap and mismatch. However, the alternative to regional analyses, a site by site approach, depends on immense amounts of data (Caughley, 1994; Oates, 1994). While we have the distributional data for primates, which is precisely the reason for their use here to examine the hotspot approach, we do not for many other taxa, especially tropical taxa (Harcourt, 1995; Howard et al., 1998). Although none of the mismatches identi®ed by this study mean that the hotspot approach is useless, the analysis con®rms how careful its application needs to be, especially if it is to be applied globally (Pressey et al., 1993; Pearson and Carroll, 1998; Reid, 1998; Ricketts et al., 1999). Hotspot analysis, like much biological conservation, concentrates on threatened wildlife more than the threats that make the analysis necessary in the ®rst place. The South American mismatch between hotspots of diversity and hotspots of threatened species emphasises the necessity of bringing together our knowledge of biodiversity with our knowledge and understanding of the intensity and nature of threats to that biodiversity. The conservation biologist's hotspot approach could surely bene®t from more analyses that add hotspots of human threat to the analysis of diversity hotspots in order to, for example, more exactly distinguish safe, diverse sites (where conservation will be cheap) from


A.H. Harcourt / Biological Conservation 93 (2000) 163±175

highly threatened, diverse sites (where conservation is expensive but vital). For instance, Dobson et al.'s (1997) hotspot analysis indicates that in the USA, the latter situation obtains, because hotspots of overall density of endangered species overlap signi®cantly with hotspots of agricultural output. Caughley (1994) has suggested that much conservation biology su€ers because ecology, unlike population biology, has so few general theories on which conservation management can be based. Consequently, he suggests, conservation often proceeds on an inecient case by case basis. These results from a detailed analysis of one taxon that shows such unexplained variability across continents and taxa in overlap of hotspots are a good illustration of Caughley's point. We could not a priori predict what hotspots would have overlapped on what continents by how much. Thus we proceed only empirically. Biogeography is the obvious theoretical discipline to bring to hotspot analysis if we want general predictability. While the original Pleistocene refugium hypothesis (Ha€er, 1969) promised a general theory behind the hotspot approach, forest refugia are proving dicult to demonstrate (e.g. Beven et al., 1984; Oates, 1988; Colyn et al., 1991; Colinvaux et al., 1996; Jolly et al., 1997). However, various other areas of biogeography seem directly applicable, such as refugium theory in general, vicariance biogeography, and the biogeography of abundance (e.g. Vermeij, 1978; Brundin, 1988; Lynch, 1988; Hengeveld, 1990; Gaston, 1994; Brown, 1995; Rosenzweig, 1995). If analysis of single orders has anything to contribute to conservation, the next step is to intensify application of these areas of biogeography to explaining the distributions of the members of the order, primates. Acknowledgements The author wishes to thank Mark Byars and especially Sean Parks for producing the hotspot maps, and Guy Cowlishaw, John Oates, Sean Parks, Walter Reid, Kelly Stewart and the journal's referees for comments that considerably improved the paper. References Beven, S., Conner, E.F., Beven, K., 1984. Avian biogeography in the Amazon basin and the biological model of diversi®cation. J. Biogeog 11, 383±399. Brandon-Jones, D., 1996. The Asian Colobinae (Mammalia: Cercopithecidae) as indicators of Quaternary climatic change. Biol. J. Linn. Soc 59, 327±350. Brown, J.H., 1995. Macroecology. University of Chicago Press, Chicago. Browne, J., 1983. The Secular Ark. Studies in the History of Biogeography. Yale University Press, New Haven, Connecticut. Brundin, L.Z., 1988. Phylogenetic biogeography. In: Myers, A.A., Giller, P.S. (Eds.), Analytical Biogeography. Chapman and Hall, London, pp. 343±369.

Caughley, G., 1994. Directions in conservation biology. J. Anim. Ecol 63, 215±244. Clutton-Brock, T.H., Harvey, P.H., 1977. Primate ecology and social organization. J. Zool., Lond 183, 1±39. Colinvaux, P.A., 1996. de Oliveira, P. E., Moreno, J.E., Miller, M.C. and Bush, M.B. A long pollen record from lowland Amazonia: forest and cooling in glacial times. Science 274, 85±88. Colyn, M., Gautier-Hion, A., Verheyen, W., 1991. A re-appraisal of palaeoenvironmental history in Central Africa: evidence for a major ¯uvial refuge in the Zaire Basin. J. Biogeog 18, 403±407. Corbet, G.B., Hill, J.E., 1991. A World List of Mammalian Species. Oxford University Press, Oxford. Corbet, G.B., Hill, J.E., 1992. The Mammals of the Indomalayan Region: A Systematic Review. Oxford University Press, Oxford. Crozier, R.H., 1997. Preserving the information content of species: genetic diversity, phylogeny, and conservation worth. Ann. Rev. Ecol. Syst 28, 243±268. Dingle, H., Carroll, S.P., Loye, J.E., 1997. Conservation, behavior, and 99% of the world's biodiversity: is our ignorance really bliss? In: Clemmons, J.R., Buchholz, R. (Eds.), Behavioral Approaches to Conservation in the Wild. Cambridge University Press, New York, pp. 72±92. Dobson, A.P., Rodriguez, J.P., Roberts, W.M., Wilcove, D.S., 1997. Geographic distribution of endangered species in the United States. Science 275, 550±553. ESRI Inc., 1998a. ARC/INFO, 7.1.2. Environmental Systems Research Institute, Inc., Redlands, California. ESRI Inc., 1998b. ArcView GIS, 3.1. Environmental Systems Research Institute, Inc., Redlands, California. Ford, S.M., 1994. Taxonomy and distribution of the owl monkey. In: Baer, J.F., Weller, R.E., Kakoma, I. (Eds.), Aotus: The Owl Monkey. Academic Press, San Diego, pp. 1±57. Gaston, K.J., 1994. Rarity. Chapman and Hall, London. Groves, C.P., 1993. Order Primates. In: Wilson, D.E., Reeder, D.M. (Eds.), Mammal Species of the World: A Taxonomic and Geographic Reference. Smithsonian Institution Press, Washington, D.C, pp. 243±277. Hacker, J.E., Cowlishaw, G., Williams, P.H., 1998. Patterns of African primate diversity and their evaluation for the selection of conservation areas. Biol. Conserv. 84, 251±262. Ha€er, J., 1969. Speciation in Amazonian forest birds. Science 165, 131±137. Hannah, L., Lohse, D., Hutchinson, C., Carr, J.L., Lankerani, A., 1994. A preliminary inventory of human disturbance of world ecosystems. Ambio 23, 246±250. Harcourt, A.H., 1995. Population viability estimates: theory and practice for a wild gorilla population. Conserv. Biol 9, 134±142. Harcourt, A.H., 1996. Is the gorilla a threatened species? How should we judge? Biol. Conserv 75, 165±176. Harcourt, C.S., Thornback, J., 1990. Lemurs of Madagascar and the Comoros. IUCN Ð The World Conservation Union, Gland. Hengeveld, R., 1990. Dynamic Biogeography. Cambridge, University Press, Cambridge. Hershkovitz, P., 1987a. The taxonomy of South American sakis, genus Pithecia (Cebidae, Platyrrhini): a preliminary report and critical review with the description of a new species and a new subspecies. Am. J. Primatol 12, 387±468. Hershkovitz, P., 1987b. Uacaries, New World monkeys of the genus Cacajao (Cebidae, Platyrrhini): a preliminary taxonomic review with the description of a new subspecies. Am. J. Primatol 12, 1±53. Hershkovitz, P., 1990. Titis, New World Monkeys of the genus Callicebus (Cebidae, Plattyrrhini): a preliminary taxonomic review. Fieldiana Zoology 55, 1±109. Howard, P.C., Viskanic, P., Davenport, T.R.B., Kigenyi, F.W., Baltzer, M., Dickinson, C.J., Lwanga, J.S., Matthews, R.A., Balmford, A., 1998. Complementarity and the use of indicator groups for reserve selection in Uganda. Nature 394, 472±475.

A.H. Harcourt / Biological Conservation 93 (2000) 163±175 IUCN, 1996. 1996 IUCN Red List of Threatened Animals. IUCN, Gland, Switzerland. Jablonski, D., 1995. Extinctions in the fossil record. In: Lawton, J.H., May, R.M. (Eds.), Extinction Rates. Oxford University Press, Oxford, pp. 45±54. Janzen, D.H., 1988. Tropical dry forests. The most endangered major tropical ecosystem. In: Wilson, E.O., Peter, F.M. (Eds.), Biodiversity. National Academy Press, Washington, D.C, pp. 130±137. Jernvall, J., Wright, P.C., 1998. Diversity components of impending primate extinctions. Proc. Nat. Acad. Sci., USA 95, 11279±11283. Jolly, D., Taylor, D., Marchant, R., Hamilton, A., Bonne®lle, R., Buchet, G., Riollet, G., 1997. Vegetation dynamics in central Africa since 18,000 yr BP: pollen records from the interlacustrine highlands of Burundi, Rwanda and western Uganda. J. Biogeog 24, 495±512. Kappeler, P.M., Heymann, E.W., 1996. Nonconvergence in the evolution of primate life history and socio-ecology. Biol. J. Linn. Soc 59, 297±326. Kinzey, W.G. (Ed.), 1997. New World Primates. Ecology, Evolution and Behavior. Aldine de Gruiter, New York. Leader-Williams, N., Albon, S.D., 1988. Allocation of resources for conservation. Nature 336, 533±535. Lernould, J.-M., 1988. Classi®cation and geographical distribution of guenons: a review. In: Gautier-Hion, A., BourlieÁre, F., Gautier, J.P., Kingdon, J. (Eds.), A Primate Radiation: Evolutionary Biology of the African Guenons. Cambridge University Press, Cambridge, pp. 54±78. Lynch, L.Z., 1988. Refugia. In: Myers, A.A., Giller, P.S. (Eds.), Analytical Biogeography. Chapman and Hall, London, pp. 311±342. Manne, L.L., Brooks, T.M., Pimm, S.L., 1999. Relative risk of extinction of passerine birds on continents and islands. Nature 399, 258±261. Maurer, B.A., Brown, J.H., Rusler, R.D., 1992. The micro and macro in body size evolution. Evolution 46, 939±953. Messier, W., Stewart, C.-B., 1997. Episodic adaptive evolution of primate lysozymes. Nature 385, 151±154. Miller, K., Allegretti, M.H., Johnson, N., Jonsson, B., 1995. Measures for conservation of biodiversity and sustainable use of its components. In: Heywood, V.H., Watson, R.T. (Eds.), Global Biodiversity Assessment. Cambridge University Press, Cambridge, pp. 915±1061. Mittermeier, R.A., Myers, N., Thomsen, J.R., 1998. da Fonseca, G. A.B. and Olivieri, S. Biodiverity hotspots and major tropical wilderness areas: approaches to setting conservation priorities. Conserv. Biol 12, 516±520. Mittermeier, R.A., Tattersall, I., Konstant, W.R., Meyers, D.M., Mast, R.B., Nash, S.D., 1994. Lemurs of Madagascar. Conservation International, Washington, D.C. Mrosovsky, N., 1997. IUCN's credibility critically endangered. Nature 389, 436. Myers, N., 1984. The Primary Source. Tropical Forests and Our Future. W.W. Norton and Co, New York. Myers, N., 1988. Threatened biotas: hotspots in tropical forests. The Environmentalist 8, 178±208. Nash, L.T., Bearder, S.K., Olsen, T.R., 1989. Synopsis of Galago species characteristics. Int. J. Primatol 10, 57±80. Niemitz, C., 1984. Taxonomy and distribution of the genus Tarsius Storr, 1780. In: Niemitz, C. (Ed.), Biology of Tarsiers. Gustav Fischer, Stuttgart, pp. 1±16. Oates, J.F., 1988. The distribution of Cercopithecus monkeys in West African forests. In: Gautier-Hion, A., BourlieÁre, F., Gautier, J.-P. (Eds.), A Primate Radiation: Evolutionary Biology of the African Guenons. Cambridge University Press, Cambridge, pp. 79±103. Oates, J.F., 1994. Africa's primates in 1992: conservation issues and options. Am. J. Primatol 34, 61±71.


Oates, J.F., Davies, A.G., Delson, E., 1994. The diversity of living colobines. In: Davies, A.G., Oates, J.F. (Eds.), Colobine Monkeys. Their Ecology, Behaviour and Evolution. Cambridge University Press, Cambridge, pp. 45±73. Olson, D.M., Dinerstein, E., 1998. The Global 2000: a representation approach to conserving the earth's most biologically valuable ecoregions. Conserv. Biol 12, 502±515. Pearson, D.L., Carroll, S.S., 1998. Global patterns of species richness: spatial models for conservation planning using bioindicator and precipitation data. Conserv. Biol 12, 809±821. Prendergast, J.R., Quinn, R.M., Lawton, J.H., Eversham, B.C., Gibbons, D.W., 1993. Rare species, the coincidence of diversity hotspots and conservation strategies. Nature 365, 335±337. Pressey, R.L., Humphries, C.J., Margules, C.R., Vane-Wright, R.I., Williams, P.H., 1993. Beyond opportunism: key principles for systematic Reserve selection. Trends Ecol. Evol 8, 124±128. Reed, K.E., Fleagle, J.G., 1995. Geographic and climatic control of primate diversity. Proc. Nat. Acad. Sci., USA 92, 7874±7876. Reid, W.V., 1998. Biodiversity hotspots. Trends Ecol. Evol 13, 275±280. Ricketts, T.H., Dinerstein, E., Olson, D.M., Loucks, C., 1999. Who's were in North America? BioScience 49, 369±381. Rosenzweig, M.L., 1995. Species Diversity in Space and Time. Cambridge University Press, Cambridge. Rowe, N., 1996. The Pictorial Guide to the Living Primates. Pogonias Press, East Hampton, NY. Rylands, A.B., Coimbra-Filho, A.F., Mitermeier, R.A., 1993. Systematics, geographic distribution, and some notes on the conservation status of the Callitrichidae. In: Rylands, A.B. (Ed.), Marmosets and Tamarins. Systematics, Behaviour, and Ecology. Oxford University Press, Oxford, pp. 11±77. Scott, J.M., Anderson, H., Davis, F., Caicoo, S., Csuti, B., Edwards, T.C.J., Noss, R., Ulliman, J., Groves, C., Wright, R.G., 1993. Gap analysis a geographic approach to protection of biological diversity. Wildl. Monogr 123, 1±41. Shepherd, U.L., 1998. A comparison of species diversity and morphological diversity across the North American latitudinal gradient. J. Biogeog 25, 19±29. Terborgh, J., Schaik, C.P.v., 1987. Convergence vs nonconvergence in primate communities. In: Gee, J.H.R., Giller, P.S. (Eds.), Organization of Communities Past and Present. Blackwell Scienti®c Publications, Oxford, pp. 185±203. Terborgh, J., Winter, B., 1983. A method for siting parks and reserves with special reference to Columbia and Ecuador. Biol. Conserv 27, 45±58. van Jaarsveld, A.S., Freitag, S., Chown, S.L., Muller, C., Koch, S., Hull, H., Bellamy, C., KruÈger, M., EndroÈody-Younga, S., Mansell, M.M., Scholtz, C.H, 1998. Biodiversity assessment and conservation strategies. Science 279, 2106±2108. Vermeij, G.J., 1978. Biogeography and Adaptation. Harvard University Press, Cambridge, MA. Wallace, A.R., 1876. The Geographical Distribution of Animals. Macmillan, London. Williams, P.H., Gaston, K.J., 1994. Measuring more of biodiversity: can higher-taxon richness predict wholesale species richness Biol? Conserv 67, 211±217. Williams, P.H., Gaston, K.J., Humphries, C.J., 1997. Mapping biodiversity value worldwide: combining higher-taxon richness from different groups. Proc. Roy. Soc. Lond. B 264, 141±148. Williams, P.H., Humphries, C.J., 1997. Comparing character diversity among biotas. In: Gaston, K.J. (Ed.), Biodiversity. A Biology of Numbers and Di€erences. Blackwell Science, Oxford, pp. 54±76. Wolfheim, J.H., 1983. Primates of the World: Distribution, Abundance and Conservation. University of Washington Press, Seattle.