For Review Only - Simon Fraser Universityamooers/papers/Gudde_etal_DD.pdf · 2012. 7. 21. · For...

For Review Only

Imperiled phylogenetic endemism of Malagasy

lemuriformes

Journal: Diversity and Distributions

Manuscript ID: DDI-2012-0087.R1

Manuscript Type: Biodiversity Research

Date Submitted by the Author: n/a

Complete List of Authors: Gudde, Renske; University of Utrecht, Biological Sciences Joy, Jeff; Simon Fraser University, Biological Sciences Mooers, Arne; Simon Fraser Unversity, Biological Sciences;

Keywords: Conservation, phylogeny, lemurs, endemism, risk, Madagascar

Diversity and Distributions


For Review Only

Imperiled phylogenetic endemism

1

1

2

Imperiled phylogenetic endemism of Malagasy lemuriformes 3

4

Renske M. Gudde1,2,5,6

, Jeffrey B. Joy2,3,4,5

, and Arne O. Mooers2,3,4,5*

5

6

1 Behavioural Biology, Department of Biology, Utrecht University, Utrecht, The Netherlands 7

2.IRMACS, Simon Fraser University, Burnaby, B.C. Canada 8

3.Department of Biological Sciences, Simon Fraser University, Burnaby, B.C. Canada 9

4 Human Evolutionary Studies Program, Simon Fraser University, Burnaby, B.C. Canada 10

5 All three authors contributed equally 11

6 Current address: Department of Biological Sciences, University of Hull, Kingston-upon-Hull, 12

UK 13

14

Running head: mapping lemur phylogeny 15

16

Keywords: Conservation, phylogeny, endemism, risk, lemurs, Madagascar 17

18

Word count: 4994 main text, 6723 all in. 19

20

*Author for correspondence 21

Email: [email protected] 22

23

Page 1 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


2

Abstract 24

Aim: To highlight where in Madagascar the phylogenetically and spatially rare lemur species at 25

risk of extinction are concentrated. 26

Location: Madagascar 27

Methods: Phylogenetic endemism (PE) is a combined measure for apportioning a phylogenetic 28

tree across a landscape used to identify those geographic locations that contain spatially rare 29

phylogenetic diversity (Rosauer et al., 2009). We present a simple extension (imperiled 30

phylogenetic endemism) that scales this phylogenetic diversity by the probability of its loss to 31

extinction. We apply these measures to a composite phylogeny of all confirmed Malagasy 32

lemuriform species using International Union for Conservation of Nature (IUCN) extent of 33

occurrence and threat status data. 34

Results: We find that, because non-imperiled species are scattered about the lemuriform tree, 35

interior branches of the tree are still quite secure: this also means that areas of high phylogenetic 36

endemism for Madagascar lemuriformes are often the same areas as those of high imperiled PE 37

(IPE), since both are heavily weighted by branches nearer the tips. However, although the North 38

of Madagascar holds the largest amount of spatially rare evolutionary history using both PE and 39

IPE, there are additional pockets of imperiled history in the south and west. 40

Main Conclusion: Correlations of endemism and threat status with phylogenetic isolation are 41

modest across lemurs and so are not substitutable conservation values. They might best be 42

integrated on the landscape using IPE. As illustrated here, IPE successfully highlights areas 43

containing species which are at once threatened with extinction and that are phylogenetically and 44

spatially rare. 45

46

47

Introduction 48

Endemic species, species unique to a specific geographic location, or sets of such species, have 49

often been used to assign conservation priority to one geographic locale over another (Myers et 50

al., 2000). Phylogenetic diversity (PD) identifies the amount of genetic diversity represented by 51

sets of species found at different locales (Faith, 1992), and this has been presented as a 52

complementary method for focusing conservation efforts on the landscape: this genetic diversity 53

is often considered a proxy for combined feature diversity (Faith, 1992, Forest et al., 2007) or 54

Page 2 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


3

ecological breadth (Cadotte et al., 2008). Recently, Rosauer et al. (2009) combined endemism 55

and PD into a metric called phylogenetic endemism (PE), which identifies geographic areas that 56

contain the most spatially restricted phylogenetic diversity. 57

58

Practically, PE is calculated by dividing all the branches of a phylogenetic tree and apportioning 59

them onto a landscape grid such that each branch length (Lc) is divided equally among the grid 60

cells in which that branch is represented (Rc). The sum of these partial branch lengths in a 61

particular cell is its total PE (Rosauer et al., 2009, Mooers and Redding, 2009). Rosauer et al. 62

(2009) presented the formula of PE as follows: 63

phylogenetic endemism = Lc

Rcc∈C{ }

∑ (1)

64

A grid cell containing a highly endemic species not represented in other grid cells will be 65

allocated the entire branch on the tree leading to that species, which will be longer to the extent 66

that the species has no close relatives, whereas a more common species will share its unique 67

branch among all the grid cells in which it occurs. Internal branches on the tree (those leading to 68

clades rather than to species) are divided up among all the grid cells in which their clade 69

members are found. Thus, PE measures areas that contain the species that are both 70

phylogenetically distinct and spatially rare. 71

72

A third and common way in which species and areas are prioritized for conservation is risk of 73

exinction, such that more imperiled species are given higher priority (see, e.g., Purvis et al., 2000; 74

Rodrigues et al., 2006). This is especially important for landscapes with endemic species, as they 75

are irreplaceable (Brooks et al., 2006). Here, we introduce an extension of PE – imperiled 76

phylogenetic endemism (IPE) - that combines PE with measures of species threat status. We test 77

the application of this metric using a new composite phylogeny of all confirmed species of 78

Malagasy lemuriform primates (n=67). 79

80

The Malagasy lemuriformes provide a valuable system for testing and applying PE and IPE: they 81

are restricted to a relatively small area (the island of Madagascar), their geographic ranges are 82

fairly well-known, and a variety of gene sequence data are available for most species (n = 57 of 83

67 confirmed species). Furthermore, many lemur species are threatened with extinction: 51 84

Page 3 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


4

confirmed lemur species have been assigned a threat status by the IUCN, and 23 are either 85

endangered or critically endangered. In addition, conservation planning is urgently needed for 86

the lemurs of Madagascar more generally: many species are threatened by subsistence hunting 87

(Lehman and Wright, 2000), logging, and habitat degradation due to agricultural activities 88

(Johnson and Overdorff, 1999) throughout their ranges. Only 10-20% of the original forest of 89

Madagascar used by many lemur species remains, and that in small and fragmented areas 90

(Lehman et al., 2006; Ganzhorn, 2001; Whitmore, 2000, Harper et al., 2007). 91

92

Below we report the first complete composite phylogenetic tree of Malagasy lemuriform primates 93

based on sequence data and use it to calculate and compare the quantities of phylogenetic 94

endemism and imperiled phylogenetic endemism on the landscape. Because species-level range 95

maps for lemurs need refining (Harcourt et al., 1990), we do not aim to present a prioritization for 96

Madagascar reserves. Rather, we quantify the effects of novel and existing conservation methods 97

on identifying areas of conservation worth, in the hopes that better maps can be generated soon 98

and that managers will subsequently consider these methods. 99

100

Methods 101

Lemuriform Taxonomy and Dataset 102

The IUCN Red List (2008; http://www.iucnredlist.org/apps/redlist/search, accessed 2-2-2011) 103

comprises 93 lemur species; however, 26 of these are delineated solely through consideration of 104

genetic distance or geographic separation of ranges and the IUCN calls for confirmation of their 105

species status (IUCN Red List, 2008; see also Tattersall, 2007). Thus there are 67 confirmed 106

lemur species. Genetic data for 57 of these 67 species were available on GenBank. We 107

reconstructed phylogenetic relationships among Malagasy Lemuriformes from coding and non-108

coding regions of 12 mitochondrial genes (Cytochrome B, Cytochrome Oxidase Subunit II, 109

Cytochrome Oxidase Subunit III, tRNA-Gly, ND3, tRNA-Arg, ND4, tRNA-His, tRNA-Ser, 110

tRNA-Leu, D-loop, and 12S) and 6 nuclear genes and introns (Short-wave Sensitive type 1 111

Opsin, Fibrinogen Alpha, Adenosine Receptor A3, ENO1, Interphotoreceptor Retinoid Binding 112

Protein, and von Willebrand Factor), all obtained from Genbank. We aligned each gene 113

separately, employing the local alignment tool MAFFT (Kathoh et al., 2009). We then inspected 114

the alignment of each gene by eye using Se-Al (Rambaut, 1996). Aligned sequences for each 115

Page 4 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


5

gene for each species were then concatenated into a supermatrix using the sequence handling 116

functions found in the R package APE (Paradis, 2006), with the remainder of the matrix filled 117

with missing data. 118

119

Phylogenetic Analyses 120

We employed both Maximum Likelihood (ML) and Bayesian methods of phylogenetic tree 121

reconstruction. For ML and Bayesian analyses we employed codon position (CP) models 122

(Shapiro et al., 2006) for protein coding sequences and estimated the best fit model of molecular 123

evolution for intron sequences and non-protein coding DNA using MrModeltest 3.04 (Nylander, 124

2004). ML analyses were conducted using RAxML version 7.04 (Stamatakis, 2006) employing 125

the combined ML search and bootstrapping function implemented on the SFU IRMACS 126

computing cluster. We conducted Bayesian analyses using MrBayes version 3.1.2 (Ronquist and 127

Huelsenbeck, 2003). Two runs employing four chains (three heated and one cold) were run for 10 128

x 106 generations, and trees were sampled every 1000 generations. We assessed convergence 129

using the standard deviation of the split frequencies between runs, and graphically using the 130

program Tracer (Rambaut and Drummond, 2007) and AWTY (Nylander et al., 2008). Bayesian 131

posterior probabilities and ML bootstrap values were utilized to assess support for recovered 132

nodes. We then rate-smoothed the consensus tree from the posterior distribution using penalized 133

likelihood and a smoothing parameter (lambda) value of 1 as implemented in APE. 134

135

Completing the tree 136

We added ten species that are recognized by the IUCN but for which we lacked sequence data to 137

our ultrametric penalized likelihood tree based upon their taxonomy (see also Day et al., 2008; 138

Lanfear and Bromham, 2011). We first used the relevant modern taxonomic treatments to place 139

each species on the tree next to its sister species: Phaner electromontis with P. furcifer (Groves 140

and Tattersall, 1991); P. parienti with P. furcifer (Groves and Tattersall, 1991); Avahi betsileo, A. 141

meridionalis, A. peyrierasi, and A. ramanantsoavani with A. laniger (Andrianantompohavana et 142

al., 2007); Cheirogaleus adipicaudatus with C. medius (Groves, 2000); C. minusculus and C. 143

ravus with C. major (Groves, 2000); and Eulemur rufifrons with E. rufus (Mittermeier et al., 144

2008). Given that all genera (complete and incomplete) were monophyletic on our tree, assuming 145

monophyly for lemuriform genera with missing species seems justifiable. We then assigned 146

Page 5 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


6

branch lengths to these new species by first assuming a simple pure-birth model of evolution for 147

the new subclade and then estimating the species age given the ML estimate of the local 148

diversification rate l for a subtree with n tips and stem age t, which is ln(n)/t (see, e.g. Magallon 149

and Sanderson, 2001). The estimated species age is 1/(2*l) (Steel and Mooers, 2010) (see results 150

below). Thus, our final composite tree contains all 67 species recognized by the IUCN. An 151

alternative approach, which ultrametricizes the posterior distribution of trees, and then adds in the 152

10 species using the techniques of Kuhn et al. (2011) yields indistinguishable trees. This full 153

distribution of trees (which may be of use for a wider range of analyses, has been deposited on 154

TreeBase (http://purl.org/phylo/treebase/phylows/study/TB2:S12843). 155

156

Geo-spatial Analysis 157

The spatial data comprising the geographical range of each lemur species were taken from the 158

IUCN database for spatial data of mammals (http://www.iucnredlist.org/technical-159

documents/spatial-data, accessed 27-08-2010). We obtained range data for 66 of the 67 160

confirmed lemur species (no spatial data on Mirza zaza are available, and so it was dropped from 161

the tree). Maps for every branch in the tree (both external and internal) were generated using the 162

mapping functions of R (R Development Core Team, 2009) (packages: maps, mapdata, maptools, 163

sp, rgdal, spatial, spdep, shapefiles and mapproj) using an equirectangular projection and a grid 164

cell size of 30 arc minutes (1/4 degree) of latitude and longitude, producing a sample size of 206 165

grid cells in which at least one lemur species was found. Grid cells differ slightly in size due to 166

this projection, thus the average size of grid cells is 2920 km2 with a standard deviation of 68 167

km2. The lemur IUCN maps have a coarse resolution and are advised to be only used for very 168

large global or continental scales (Hoffmann et al., 2008). Hurlbert and Jetz (2007) suggest that 169

the grid cell size chosen for coarse range size maps should be at least 2 degrees in order to avoid 170

overestimation of species richness. However, Madagascar can be divided into 26 grid cells when 171

the size is 2 degrees. As our study is primarily heuristic we chose the grid cell size of a quarter 172

degree, resulting in 206 grid cells that contain lemurs. Our results are reported for this scale only, 173

and top-ranking grid cells were then identified on a map of natural and human used environment 174

in Madagascar developed by Bidgrain (2010). 175

176

Phylogenetic Endemism 177

Page 6 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


7

We calculated phylogenetic endemism for the Malagasy lemurs by first obtaining the length of 178

each branch in the phylogenetic tree using the tree handling functions in the Ape. Second, the 179

number of grid cells in which the branch or clade occurs was calculated using a program written 180

in C# (Map Reader, available on request) and each branch of the tree was divided equally among 181

that number. Using Excel, these branch fractions were then apportioned to each grid cell with 182

Map Reader, using the maps of each species or the union of such maps for deeper branches). The 183

resulting distribution of quantities of lemuriform evolutionary history on the landscape was then 184

visualized using heat maps (Wilkinson and Friendly, 2008) in R projected on to the map of 185

Madagascar using the original 30 arc minutes grid. We also constructed heatmaps for species 186

richness per grid cell (SR), and Weighted Endemism per grid cell, calculated by taking the 187

inverse of the number of grid cells in which a species x occurs (Qx) and summing the total for all 188

S species in a grid cell, following Kerr (1997; Crisp et al., 2001): 189

WE = Qx

−1

x=1

S

∑ (2)

190

191

Risk-Weighted Phylogenetic Endemism 192

We can extend the idea of apportioning the tree to the landscape to explicitly include the current 193

probability of loss of portions of that tree. Suppose we have a tree with E total branches, each 194

branch (or edge) e assigned a length l(e), and we assign a probability of extinction (p(ext)x) to 195

each tip x in the tree. Witting and Loeschke (1995) present a simple equation for the expected 196

total loss of phylogenetic diversity from that tree: 197

198

E(Loss) = (l(e) p(ext)x

x∈C (e )

∏e∈E

∑ ) (3) 199

200

Where E(Loss) is the expected loss, and C(e) denotes the set of tips descendent from branch e. 201

For external branches (leading to species), the expected loss is simply the length of the branch 202

(the first term in the summation) times its probability of extinction. The expected loss for an 203

internal branch is its length multiplied by the probability that all the species it subtends go extinct 204

(hence the product term). The expected total loss is then the sum of the expected loss terms for 205

each branch. Magnuson-Ford et al. (2010) make explicit that this quantity is closely related to the 206

expression for PD that is expected to remain in the future (E(PD); Faller et al., 2008) such that 207

Page 7 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


8

E(PD) = PD – E(Loss). The concept of expected loss is also behind the ‘edge of existence’ 208

conservation programme, which highlights threatened species representing unique evolutionary 209

history (see Redding and Mooers, 2006; Isaac et al., 2007, Collen et al., 2011). 210

211

Assuming one can assign p(ext) to the tips of the tree, equation (2) can be used to produce a tree 212

whose branch lengths are in units of expected loss (see Figure 1). If we apply the PE algorithm to 213

this transformed tree, we quantify where on the landscape at-risk evolutionary history is 214

concentrated. We call this mapped quantity ‘imperiled phylogenetic endemism’ or IPE. 215

Following Table 1 from Mooers et al. (2008) and drawing on Redding and Mooers (2006) and 216

Isaac et al. (2007), we applied two transformations to IUCN categories to produce rough p(ext) 217

values (see our Table 1). Both use a nominal 100-year window and sets p(ext) for vulnerable 218

species = 0.1, with the IUCN transformation having a much steeper slope between shifts in IUCN 219

categories and p(ext). Rather than assign a single p(ext) value to the 15 'data deficient' species, we 220

followed Magnuson-Ford et al. (2010) and used information on their ecology to estimate their 221

threat status (Table S1 contains all taxonomic and threat status information). We report results 222

from both transformations, with a focus on the IUCN transformation. 223

224

Correlations Among Measures 225

To examine how different methods of reserve rankings might differ, we compared PD, PE, IPE, 226

species richness and endemism with Pearson’s correlations as well as with linear and polynomial 227

regressions across all 206 grid cells containing lemur species using R (R Development Core 228

Team 2011). As PD, PE and IPE are (weighted) sums of branch lengths found in a grid cell we 229

expect them to share variation between them and with raw species richness and weighted 230

endemism (Mooers and Redding, 2009). We also considered the overlap between the measures 231

for the top 20 cells – i.e. for those cells with greatest conservation worth. 232

233

Results 234

Sequence Dataset 235

Our sequence dataset consisted of 302 sequences for 57 lemur taxa comprising a total of 13,127 236

positions (Table S2, Supplementary information). The matrix is 39.4% complete. Of the 13,127 237

positions 4062 characters were variable and 3062 characters were parsimony informative. Taxon 238

Page 8 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


9

and character sampling was heterogeneous among higher-level lemurs. We included the pygmy 239

slow loris (Nycticebus pygmaeus) and thick-tailed galago (Otolemur crassicaudatus) as outgroup 240

taxa. Genbank accession numbers for all genes included in our analyses are listed in Table S2, 241

Supporting information. 242

243

Phylogeny of Malagasy Lemuriformes 244

We summarized trees sampled during the last 2 million generations of mcmc sampling. Both ML 245

(RAxML) and Bayesian (MrBayes 3.1.2) methods converged on nearly identical topologies. 246

Similarly, with few exceptions, support for familial and generic relationships as judged by both 247

ML (RAxML) bootstrapping and Bayesian posterior clade probabilities were robust throughout 248

the topology (Figure 2). No well-supported nodes differed among analyses. The composite 249

ultrametric phylogeny of Malagasy lemuriform primates that includes species lacking sequence 250

data that were added to the tree is presented in Figure 3. Importantly, the average branch lengths 251

of internal and external branches in the complete portions of the tree are very similar, supporting 252

our use of a Yule model for placing the missing species on the tree. Based on our results the 253

Malagasy lemuriformes are composed of 5 major clades broadly corresponding to families. 254

Among families lemur genera were each reciprocally monophyletic, receiving Bayesian posterior 255

clade probabilities of 1.0 (RAxML bootstrap values of 100), with the exception of the placement 256

of the genus Phaner (supported with 0.97 posterior probability, RAxML bootstrap value of 73). 257

There were few instances of poorly supported relationships within lemur genera, notably within 258

Eulemur, Propithecus, and among the Microcebus species (Figure 2). 259

Figure 4 depicts the corresponding risk-weighted tree using the IUCN transformation for 260

probabilities of extinction; Figure S1, Supplementary information, presents the tree under both 261

the Isaac and IUCN transformations for comparison. 262

263

Phylogenetic Endemism 264

The heat map for PD is given in Figure 5D and PE is depicted in Figure 5E. The 20 grid cells 265

with the highest PE values are primarily found in the north and the east of the Island of 266

Madagascar. Interestingly, though PE does show significant clustering, as might be expected 267

(Join Count test on the top 20 cells, Z=31, P<0.001), some of the highest ranked grid cells are 268

dispersed rather than clustered (e.g. the grid cell with the highest and third highest PE can be 269

Page 9 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


10

found in the east of Madagascar, whereas the second highest PE level is found in the north). The 270

20 highest ranking grid cells all contain differing amounts of fragmented areas of forest (Bidgain, 271

2010). Half of the top 20 PE grid cells contain national and special reserves, and 6 of the 20 272

contain national parks. Forest in Madagascar is concentrated in the north, east, and along the west 273

coast (Figure 5A); though relatively species-rich (Figure 5B), the 274

west does not seem to harbour exceptional amounts of lemur PE. 275

276

Imperiled Phylogenetic Endemism 277

Figure 5F presents the heat maps of the risk-weighted lemur trees mapped onto the landscape 278

using the IPE algorithm and the IUCN transformation, and the corresponding map for the IPE 279

algorithm and the “Isaac” transformation can be found in Figure S2. Eight of the top 20 IPEIUCN 280

grid cells are found in the north of Madagascar, with the remainder scattered in the east (n = 11) 281

and west (n = 1). The top 20 IPEIUCN grid cells all contain forested habitat, 2 of these 20 grid cells 282

also contain conservation reserves, and national parks occur in 9 grid cells (Bidgrain, 2010). 283

284

Correlations among measures 285

Heatmaps of species richness (SR), weighted endemism (WE), phylogenetic diversity (PD), 286

phylogenetic endemism (PE), and Imperiled PE (IPEIUCN) are presented in Figure 5 (and IPEIsaac 287

is presented in Figure S2, Supplementary material). Pairwise rank correlation coefficients across 288

all the measures are presented in Table 2, and pairwise scatterplots are presented in Figures S3 289

and S4, Supplementary material. Rank correlations were generally high to very high, especially 290

among the various phylogenetic endemism measures. The weakest relationship was between PD 291

and Weighted endemism (r = 0.5). Nonlinear models on transformed data improved the pairwise 292

fit of relationship between variables somewhat (Table S3, Supplementary material); for instance 293

R-squared for the relationship between phylogenetic endemism and species richness improves 294

from 0.67 in a simple linear model to 0.70 in a quadratic framework. Even with non-linear fits, 295

there is appreciable non-overlapping variation (on the order of 30%) between phylogenetic and 296

nonphylogenetic measures of conservation worth. 297

298

299

300

Page 10 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


11

301

However, if we confine our analyses to the top 20 cells the overlap among metrics is only 302

moderate (Table 3). Overall, maps of species richness and weighted endemism are not very 303

congruent, having only 12 cells in common. This is due to the fact that the relevant species show 304

high variance in range size. PE and IPE show mild overlap with each other (see discrepancies in 305

particular the Northwest); this is also expected because IPE incorporates an additional 306

independent variable. PE also shows only mild overlap with SR and WE. Seventeen of the top 20 307

cells are common to the three phylogenetic methods (PE, IPEIUCN and IPEIsaac), and 10 grid cells 308

are common in the top 20 across SR, WE and the three range-weighted phylogenetic methods. 309

The overlap of with PD with the other metrics is low, and only 6 grid cells are common across all 310

six metrics. 311

312

Discussion 313

Phylogeny of Malagasy Lemuriformes 314

We present the most complete, best-supported phylogeny of the Malagasy lemuriformes to date. 315

Our tree is broadly in agreement with previous phylogenies: the 5 lemur families are shown to be 316

monophyletic and Daubentonia madagascariensis diverges early as the sister to all other lemurs. 317

Our topology is completely congruent with that of Horvath et al. (2008), with stronger support 318

(posterior of 1.0 vs. 0.77) for the clade including E. rufus, E. rubriventer, Hapalemur griseus and 319

Lemur catta. While our topology and that of the consensus tree of Chatterjee et al. (2009) are not 320

fully congruent (having an NNI or Nearest Neighbour Interchange distance of 9; see Felsenstein, 321

2004), the differences are almost exclusively in areas of lower support on one or both trees. For 322

instance, differences within the Lepilemur genus have no support on the Chatterjee (2009) tree 323

(with bootstraps <50%), while the arrangement near Microcebus myoxinus has little support from 324

our data. Importantly, a Shimodaira-Hasegawa test (Shimodaira and Hasegawa, 1999) as 325

implemented in PAUP* 4.0b10 (Swofford, 2002) shows that our consensus topology and that of 326

Chatterjee et al. (2009) are not significantly different on our (more extensive) dataset (∆ 327

lnL=19.02, P=0.34). While all trees are conditioned on available data, we hope the present one 328

will serve as a basis for more evolutionary ecology and conservation work in this important and 329

charismatic group. 330

331

Page 11 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


12

Phylogenetic Endemism and Risk-weighted Phylogenetic Endemism 332

Similarly to PE, IPE can assess rare biodiversity consistently and independently of politically and 333

biologically defined regions (Rosauer et al., 2009). IPE adds to PE the capacity to further identify 334

locations where concentrations of phylogenetically distinct and spatially and numerically rare 335

endangered species are concentrated, which allows for more in-depth scrutiny. 336

337

For instance, the grid cell in the mid-east of Madagascar (coordinates 8, 9) ranked first for PE and 338

IPEIUCN and third for both IPEIsaac and weighted endemism contains Cheirogaleus sibreei, a 339

relatively isolated species on the tree that is Endangered and restricted to only 2 disjunct grid 340

cells. This species and its habitat might be worthy of new attention in conservation planning 341

exercises. 342

343

Correlations among measures 344

Closely related species of lemurs tend to be broadly allopatric, and species ages are generally 345

similar on the tree (with the exception of the broadly-distributed D. madagascarensis, and Indri 346

and Phaner species). This drives high correlations between phylogenetic and non-phylogenetic 347

metrics generally. For instance, weighted endemism is concentrated in the north of Madagascar. 348

The highest ranked grid cell for WE harbours 12 species, 8 of which have range sizes of 8 grid 349

cells or fewer (the average number of grid cells is 21, while the median is 8). This high level of 350

weighted endemism also leads to a high rank for PE (2nd), and IPEIUCN (5th) and IPEIsaac (1st). 351

The exception here is the phylogenetic metric PD, which shows a weak correlation with WE (r = 352

0.50). Given the strong relationship between PD and SR, this shows the strong influence of range 353

size on PE. 354

355

Focussing on the newer phylogenetic measures, external branches contribute most to both PE and 356

IPE. Thus, areas of high phylogenetic endemism should be congruent with those of high 357

imperiled PE. This congruence is increased for Madagascar Lemuriformes because species with 358

low current extinction probability are distributed throughout our phylogeny (see also Magnuson-359

Ford et al., 2010), safeguarding internal branches and so down-weighting them in IPE 360

calculations (indeed, most internal branches are of negligible length under the IPE transformation 361

(Figure 3 and Figure S1, Supplemental information) attesting to their small chance of being lost. 362

Page 12 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


13

Application of IPE to taxa with a more uneven distribution of extinction probability (specifically 363

clustering of high extinction risk within small clades) is likely to lead to internal branches that are 364

less secure and possibly even more mismatch between PE and IPE. 365

366

What mismatches there are between PE and IPE are driven by the interaction among range size, 367

branch length and threat status. So, grid cells containing phylogenetically distinct and spatially 368

rare species contribute to high PE scores. If these species are relatively secure, however, then 369

they will contribute little to IPE. For example, the 19th

highest rank grid cell for PE (coordinates 370

9,5) is ranked 58th

under IPEIUCN. The distinct and small-range Lepilemur jamesi, which 371

contributes to the high rank for PE there, has low relative extinction probability, and so this grid 372

cell is not flagged as a high-priority area. 373

374

PE correlates more strongly with IPEIUCN than with IPEIsaac. This is expected, because the IUCN 375

weighting scheme ranges over four orders of magnitude vs. only two under the Isaac et al. 376

scheme (note the different scales for the two trees in Figure S1, Supplemental information). 377

However, and somewhat counter-intuitively, the more extreme IUCN weighting scheme 378

increases the correspondence between imperiled phylogenetic endemism and species richness: 379

grid cells with high species richness have a higher probability of containing at least one 380

endangered or critically endangered species. 381

382

While the risk-weighted trees using the two transformations are very congruent there are a few 383

clear outliers: the Varecia clade, Propithecus clade, and Prolemur simus are all endangered or 384

critically endangered, resulting in higher relative weights under the IUCN than under the Isaac 385

transformation. On the other hand, D. madagascariensis has a very long external branch, but is 386

listed only as Near Threatened, and so has a lower relative extinction probability for IPEIUCN than 387

for IPEIsaac. 388

389

Given the strong impact of external branch length and range size on phylogenetic endemism 390

scores, it is instructive to consider the relationships among external branch length, range size 391

(measured as number of grid cells occupied), and p(ext) across species (Figure S5, 392

Supplementary material). The correlation across all 206 grid cells is significant for ln(range size) 393

Page 13 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


14

and ln(p(ext)): r = -0.62 and -0.63 (IUCN and Isaac transformations respectively). However, 394

more than 35% of the variation in p(ext) is independent of range size. For the other two measures, 395

there is no correlation at all: ln(external branch length) and log(range size): r = 0.12; ln(external 396

branch length) vs. ln(p(ext)): r =-0.17. It is this independence that suggests that imperiled 397

phylogenetic endemism may be a useful method to highlight the areas that contain the rarest 398

species overall. 399

400

Conclusion 401

Most conservation metrics, including those used here (weighted endemism, species richness, 402

phylogenetic diversity and phylogenetic endemism) are contingent on a clear view of what 403

constitutes a lineage worth preserving. As highlighted with the Madagascar Lemuriformes (with 404

67 well-supported species, but over 90 putative species), this is a very vexing issue. Though we 405

found strong correspondence in what areas are ranked high using two different sets of extinction 406

probabilities, such probabilities are neither static nor fully quantifiable and more work is needed 407

(see also Mooers et al., 2008). Regardless, phylogenetic endemism metrics speak to a growing 408

realization that all lineages are not equal, and that geographic and evolutionary redundancy 409

should be considered explicitly in conservation alongside threat, especially in triage situations 410

(Isaac et al., 2007; Marris, 2007; Wellnitz and LeRoyPoff, 2001). Though we included PD of a 411

grid cell as a further measure of conservation worth for comparison, and phylogenetic endemism 412

has Faith's (1992) phylogenetic diversity (PD) concept at its core, the two approaches are distinct. 413

Faith’s PD measure was designed with marginal gains in mind (see, e.g. Forest et al., 2007); 414

Rosauer et al. (2009) were clear that PE was not designed with this goal. For instance, protecting 415

the top ranked areas in Figure 4 would not necessarily protect the maximum amount of lemur PD 416

overall (see also Faith 2008). Like species richness and weighted endemism, (I)PE must be 417

treated as an additional point measure of biodiversity value on the landscape. Given that it does 418

include endemism and edge lengths explicitly, it would be interesting to test its ability to capture 419

cumulative PD. Regardless, phylogenetic endemism may be a useful tool for helping identify 420

potential areas for conservation reserves, especially when good spatial and phylogenetic data are 421

available. PE and IPE would serve as useful layers when prioritizing areas for conservation 422

action, especially in a planning exercise that mapped and combined metrics of interest on the 423

landscape first, and added planning boundaries later (Rosauer et al., 2009). 424

Page 14 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


15

425

Acknowledgements 426

We are especially grateful to Tom Sanders for assistance with computer programming and 427

designing the C+ program that divided branches among grid cells based on occurrence maps. We 428

are grateful to Simon Reader for facilitating this collaboration, to Walter Jetz and the fab-lab at 429

SFU for input, and to Walter Jetz, Dan Rosauer, Karen Magnuson-Ford, Simon Reader, Tom 430

Sanders, Dan Faith and two anonymous reviewers for commenting on previous versions of this 431

manuscript. The IRMACS centre provided world class facilities and computing resources. This 432

work was supported by an NSERC Discovery grant to AOM. 433

434

References 435

Andriantompohavana, R., Lei, R., Zaonarivelo, J.R., Engberg, S.E., Nalanirina, G., McGuire, 436

S.M., Shore, G.D., Andrianasolo, J., Herrington, K., Brenneman, P.A. & Louis Jr., E.E. 437

(2007) Molecular phylogeny and taxonomic revision of the woolly lemurs, genus Avahi 438

(Primates: Lemuriformes) Museum of Texas Tech University. Special Publication, 51, 439

1-59 440

Bidgrain, P. (2010) Adaption des Programmes à l’Océan Indien (APOI) classe de sixième. 441

http://svt.ocean-indien.pagesperso-orange.fr/cartes/madagascar.htm 442

Brooks, T.M., Mittermeier, R.A., da Fonseca, G.A.B., Gerlach, J., Hoffmann, M., Lamoreux, 443

J.F., Mittermeier, C.G., Pilgrim, J.D. & Rodrigues, A.S.L. (2006) Global Biodiversity 444

Conservation Priorities. Science, 313, 58-61. 445

Cadotte, M.W., Cardinale, B.J. & Oakley, T.H. (2008) Evolutionary history and the effect of 446

biodiversity on plant productivity. Proceedings of the National Academy of Sciences 447

USA, 105, 17012-17017. 448

Chatterjee, H., Ho, S.Y.W., Barnes, I. & Groves, I. (2009) Estimating phylogeny and divergence 449

times of primates using a supermatrix approach. BMC Evolutionary Biology 9, 259. 450

Collen, B., McRae, L., Deinet, S., De Palma, A., Carranza, T., Cooper, N., Loh, J. & Baillie, 451

L.E.M. (2011) Predicting how populations decline to extinction. Philosophical 452

Transcriptions of the Royal Society, 366, 2577-2586. 453

Day, J.J., Cotton, J.A. & Barraclough, T.G. (2008) Tempo and mode of diversification of Lake 454

Tanganyika cichlid fishes. PLoS One, 3, e1730. 455

Page 15 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


16

Faith, D.P. (1992) Conservation evaluation and phylogenetic diversity. Biological Conservation, 456

61, 1–10. 457

Faith, D.P. (2008) Threatened species and the potential loss of phylogenetic diversity: 458

conservation scenarios based on estimated extinction probabilities and phylogenetic risk 459

analysis. Conservation Biology, 22, 1461–1470 460

Faller, B., Pardi, F. & Steel, M. (2008) Distribution of phylogenetic diversity under random 461

extinction. Journal of Theoretical Biology, 251, 286–296. 462

Felsenstein, J. (2004) Inferring Phylogenies. Sinauer Associates, Sunderland, Massachusetts. 463

Forest, F., Grenyer, R., Rouget, M., Davies, T.J., Cowling, R.M., Faith, D.P., Balmford, A., 464

Manning, J.C., Proches, S., Van der Bank, M., Reeves, G., Hedderson, T.A.J. & 465

Savolainen, V. (2007) Preserving the evolutionary potential of floras in biodiversity 466

hotspots. Nature, 445, 757-760. 467

Ganzhorn, J.U., Malcomber, S., Andrianatoanina, O. & Goodman, S.M. (1997) Habitat 468

characteristics and lemur species richness in Madagascar. Biotropica, 29, 331-343. 469

Groves, C.P. & Tattersall, I. (1991) Geographical variation in the fork-marked lemur, Phaner 470

furcifer (Primates, Cheirogaleidae). Folia Primatologica, 56, 39-49. 471

Groves, C.P. (2000) The genus Cheirogaleus: Unrecognized biodiversity in dwarf lemurs. 472

International Journal of Primatology, 21, 943-962. 473

Harcourt, C. & Thornback, J. (1990) Lemurs of Madagascar and the Comoros. IUCN Red 474

Data Book, IUCN, Gland, Switzerland. 475

Harper, G.J., Steininger, M.K., Tucker, C.J., Juhn, D. & Hawkins, F. (2007) Fifty years of 476

deforestation and forest fragmentation in Madagascar. Environmental Conservation, 34, 477

1-9. 478

Hoffmann, M., Brooks, T.M., da Fonseca, G.A.B., Gascon, C., Hawkins, F.A., James, R.E., 479

Langhammer, P., Mittermeier, R.A., Pilgrim, J.D., Rodrigues, A.S.L. & Silva, J.M.C. 480

(2008) Conservation planning and the IUCN Red List. Endangered Species Research, 6, 481

113-125. 482

Horvath, J.E., Weisrock, D.W., Embry, S.L., Fiorentino, I., Balhoff, J.P., Kappeler, P., Wray, 483

G.A., Willard, H.F. & Yoder, A.D. (2008) Development and application of a 484

phylogenomic toolkit: Resolving the evolutionary history of Madagascar’s lemurs. 485

Genome Research, 18, 489-499. 486

Page 16 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


17

Hurlbert, A.H. & Jetz, W. (2007) Species richness, hotspots, and the scale dependence of range 487

maps in ecology and conservation. PNAS, 104, 13384-13389. 488

Isaac, N.J.B., Turvet, S.T., Collen, B., Waterman, C. & Ballie, J.E.B. (2007) Mammals on the 489

EDGE: conservation priorities based on threat and phylogeny. PLoS One, 2, e296. 490

IUCN 2010.IUCN Red List of Threatened Species.Version 2010.4. <http://www.iucnredlist.org>. 491

Downloaded on 2 February 2011. 492

Johnson, S.E. & Overdorff, D.J. (1999) Census of brown lemurs (Eulemur fulvus spp.) in 493

southeastern Madagascar: Methods testing and conservation implications. American 494

Journal of Primatology, 47, 51-60. 495

Kathoh, K., Asimenos, G. & Toh, H. (2009) Multiple alignment of DNA sequences with 496

MAFFT. Methods in Molecular Biology, 537, 39-64. 497

Kerr, J.T. (1997) Species Richness, Endemism, and the Choice of Areas for Conservation. 498

Conservation Biology, 11, 1094-1100. 499

Lanfear, R. & Bromham, L. (2011) Estimating phylogenies for species assemblages: A complete 500

phylogeny for the past and present native birds of New Zealand. Molecular 501

Phylogenetics and Evolution, 61, 958-963 502

Lehman, S.M., Ratsimbazafy, J., Rajaonson, A. & Day, S. (2006) Ecological Correlates to Lemur 503

Community Structure in Southeast Madagascar. International Journal of Primatology, 504

27, 1023-1040. 505

Magallon, S. & Sanderson, M.J. (2001) Absolute diversification rates of Angiosperm clades. 506

Evolution, 55, 1762-1780. 507

Magnuson-Ford, K., Mooers, A.O., Paquet, S.R. & Steel, M. (2010) Comparing strategies to 508

conserve evolutionary diversity. Journal of Theoretical Biology, 266, 107-116. 509

Marris, E. (2007) What to let go. Nature 450,152-155. 510

Meyers, N., Mittermeier, R.A., Mittermeier, C.G., Da Fonseca, G.A.B. & Kent, J. (2000) 511

Biodiversity hotspots for conservation priorities. Nature, 403, 853–858. 512

Mittermeieri, R.A., Konstant, W.R., Tattersall, I., Rylands, A.B., Ratsimbazafy, J., Louis Jr., E.E. 513

& Schwitzer, C. (2008) Lemur diversity in Madagascar. International Journal of 514

Primatology, 29, 1607-1656 515

Mooers, A.O. & Redding, D.W. (2009) Where the rare species are. Molecular Ecology, 18, 3955-516

3957. 517

Page 17 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


18

Nylander, J.A. (2004) MrModeltest v2. Program distributed by the author. Evolutionary Biology 518

Centre, Uppsala Univ, Sweden. 519

Nylander, J.A., Wilgenbusch, J.C., Warren, D.L. & Swofford, D.L. (2008) AWTY (are we there 520

yet?): a system for graphical exploration of MCMC convergence in Bayesian 521

phylogenetics. Bioinformatics, 24, 581-583. 522

Paradis, E. (2006) Analysis of phylogenetics and evolution with R. New York, Springer. 523

Purvis, A., Gittleman, J.L., Cowlishaw, G. & Mace, M.M. (2000) Predicting extinction risk in 524

declining species. Proceedings of the Royal Scociety B: Biological Sciences, 267, 1947-525

1952. 526

R Development Core Team. (2011) R: A language and environment for statistical computing. R 527

Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL 528

http://www.R-project.org/. 529

Rambaut, A. (1996) Se-Al: Sequence alignment editor. Available from 530

http://evolve.zoo.ox.ac.uk/. 531

Rambaut, A. & Drummond, A.J. (2007) Tracer v1.4, Available from 532

http://beast.bio.ed.ac.uk/Tracer. 533

Rodrigues, A.S.L., Pilgrim, J.D., Lamoreux, J.F., Hoffmann, M. & Brooks, T.M. (2006) The 534

value of the IUCN Red List for conservation. Trends in Ecology and Evolution, 21, 71-535

76. 536

Ronquist, F. & Huelsenbeck, J.P. (2003) MrBayes 3: Bayesian phylogenetic inference under 537

mixed models. Bioinformatics, 19, 1572-1574. 538

Rosauer, D., Laffan, S.W., Crisp, M.D., Connellans, S.C. & Cook, L.G. (2009) Phylogenetic 539

endemism: a new approach for identifying geographical concentrations of evolutionary 540

history. Molecular Ecology, 18, 4061-4072. 541

Shapiro, B., Rambaut, A. & Drummond, A.J. (2006) Choosing Appropriate Substitution Models 542

for the Phylogenetic Analysis of Protein-Coding Sequences. Molecular Biology and 543

Evolution, 23, 7-9. 544

Shimodaira, H. & Hasegawa, M. (1999) Multiple comparisons of log-likelihoods with 545

applications to phylogenetic inference. Molecular Biology and Evolution, 16, 1114-1116 546

Stamatakis, A. (2006) RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with 547

thousands of taxa and mixed models. Bioinformatics, 22, 2688-2690. 548

Page 18 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


19

Swofford, D.L. (2002) PAUP*. Phylogenetic analysis using parsimony (*and other methods). 549

Ver. 4.Sinauer Associates, Sunderland, MA. 550

Welnitz, T. & LeRoyPoff, N. (2001) Functional redundancy in heterogeneous environments: 551

implications for conservation. Ecology Letters, 4, 177-179 552

Wilkinson, L. & Friendly, M. (2009) The history of the cluster heat map. The American 553

Statistician, 63, 179-184. 554

Witting, L. & Loeschke, V. (1995) The optimization of biodiversity conservation. Biological 555

Conservation, 71, 205-207. 556

Yoder, A.D., Rasoloarison, R.M., Goodman, S.M., Irwin, J.A., Atsalis, S., Ravosa, M.J. & 557

Ganzhorn, J.U. (2000) Remarkable species diversity in Malagasy mouse lemurs 558

(primates, Microcebus). Proceedings of the National Academy of Sciences USA, 91, 559

11325-11330. 560

561

Page 19 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


20

Figure legends 562

563

Figure 1. The production of a tree whose branch lengths are in units of expected loss. Because 564

the probability of loss of internal branches is the product of the probabilities of loss of 565

all tips above it, their worth (length) decreases relative to external branches, and their 566

relative lengths can also change substantially. 567

568

Figure 2. Bayesian consensus phylogeny of 57 Malagasy lemuriform species based on 12 nuclear 569

and mitochondrial genes. Support values are as follows: * are placed at nodes receiving 570

both 100 ML bootstrap support and 1.0 Bayesian posterior probabilities, otherwise ML 571

bootstrap/Bayesian posterior probability values are reported, '_' represent values for 572

nodes receiving ML or Bayesian support values < 70. 573

574

Figure 3. Rate-smoothed composite phylogeny of Malagasy lemuriform primates based on 12 575

nuclear and mitochondrial genes. Nodal support values are as follows: * are placed at 576

nodes receiving both 100 ML bootstrap support and 1.0 Bayesian posterior probabilities, 577

otherwise ML bootstrap/Bayesian posterior probability values are reported, '_' represent 578

values for nodes receiving ML or Bayesian support values < 70. Lineages lacking 579

sequence data that were added by hand are denoted with '– a' after the species name and 580

an 'a' at the nodes. 581

582

Figure 4. The tree from figure 2 with branch lengths in units of expected loss, using the IUCN 583

weighting scheme: longer branches signify more ‘at risk’ evolutionary history (in 584

relative time units) than do shorter branches. Open triangles above branches indicate 585

branches whose lengths are very different when comparing IUCN versus Isaac 586

probabilities of extinction (see Supplemental Figure 1). 587

588

Figure 5. (Heat) maps of conservation measures with an equirectangular projection. The numbers 589

represent the rank within the top 20, with 1 being the highest rank. Red grid cells contain 590

the highest levels, white grid cells contain the lowest levels. Grid cells are quarter-591

degree squares (30 arc minutes) with the (1,1) coordinate being (43.2°W, -25°S). 592

Page 20 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


21

A) Geopolitical map of Madagascar: dark green: forest; light green: mangroves; 593

reserves: hatched; managed environments: tan. B) Species Richness - here the cells 594

marked 'A' = rank 1 (16 species); cells marked 'B' = rank 2 (14 species); cells marked 595

‘C’ = tied rank 3 (13 species); marked 'D' = tied rank 4 (12 species); C) Weighted 596

Endemism; D) Phylogenetic Diversity; E) Phylogenetic Endemism; F) Imperiled 597

Phylogenetic Endemism (IUCN transformation). 598

599

Tables 600

601

Table 1. Species extinction probabilities based on IUCN and Isaac transformations. 602

IUCN category IUCN Isaac

Least concern 0.0001 0.025

Near threatened 0.01 0.05

Vulnerable 0.1 0.1

Endangered 0.667 0.2

Critically endangered 0.999 0.4

603

Table 2. Standard correlation coefficients among conservation metrics for all grid cells containing 604

lemurs (N = 206). 605

PE IPEIUCN IPEIsaac SR WE PD1

Phylogenetic Endemism (PE) 0.96

0.93 0.89 0.82 0.72

Imperiled PE (IPEIUCN)2

0.95 0.81 0.82 0.72

Imperiled PE (IPEIsaac)2

0.88 0.80 0.82

Species Richness (SR) 0.64 0.95

Weighted Endemism (WE) 0.50

1. PD: Phylogenetic Diversity of a grid cell (Faith, 1992) 606

2. See Table 1 for transformations of IUCN categories to nominal extinction probabilities. 607

608

Table 3. Top-20 grid cell overlap among conservation metrics1. 609

PE IPEIUCN IPEIsaac SR WE PD2

Phylogenetic Endemism (PE) 90

85 65 75 50

Page 21 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only


22

Imperiled PE (IPEIUCN)3

90 70 65 55

Imperiled PE (IPEIsaac)3

80 65 55

Species Richness (SR) 60 65

Weighted Endemism (WE) 35

1. Entries are percentage of the top twenty grid cells that are common to a pair of metrics. 610

2. PD: Phylogenetic Diversity of a grid cell (Faith, 1992) 611

3. See Table 1 for transformations of IUCN categories to nominal extinction probabilities. 612

613

614

Page 22 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

The production of a tree whose branch lengths are in units of expected loss. Because the probability of loss of internal branches is the product of the probabilities of loss of all tips above it, their worth (length) decreases relative to external branches, and their relative lengths can also change substantially.

196x105mm (72 x 72 DPI)

Page 23 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

Bayesian consensus phylogeny of 57 Malagasy lemuriform species based on 12 nuclear and mitochondrial genes. Support values are as follows: * are placed at nodes receiving both 100 ML bootstrap support and 1.0 Bayesian posterior probabilities, otherwise ML bootstrap/Bayesian posterior probability values are

reported, '_' represent values for nodes receiving ML or Bayesian support values < 70. 1114x881mm (72 x 72 DPI)

Page 24 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

Rate-smoothed composite phylogeny of Malagasy lemuriform primates based on 12 nuclear and mitochondrial genes. Nodal support values are as follows: * are placed at nodes receiving both 100 ML bootstrap support and 1.0 Bayesian posterior probabilities, otherwise ML bootstrap/Bayesian posterior

probability values are reported, '_' represent values for nodes receiving ML or Bayesian support values < 70. Lineages lacking sequence data that were added by hand are denoted with '– a' after the species name and

an 'a' at the nodes. 285x228mm (300 x 300 DPI)

Page 25 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

The tree from figure 2 with branch lengths in units of expected loss, using the IUCN weighting scheme: longer branches signify more ‘at risk’ evolutionary history (in relative time units) than do shorter branches. Open triangles above branches indicate branches whose lengths are very different when comparing IUCN

versus Isaac probabilities of extinction (see Supplemental Figure 1). 833x1003mm (72 x 72 DPI)

Page 26 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

(Heat) maps of conservation measures with an equirectangular projection. The numbers represent the rank within the top 20, with 1 being the highest rank. Red grid cells contain the highest levels, white grid cells contain the lowest levels. Grid cells are quarter-degree squares (30 arc minutes) with the (1,1) coordinate

being (43.2°W, -25°S).

A) Geopolitical map of Madagascar: dark green: forest; light green: mangroves; reserves: hatched; managed environments: tan. B) Species Richness - here the cells marked 'A' = rank 1 (16 species); cells

marked 'B' = rank 2 (14 species); cells marked ‘C’ = tied rank 3 (13 species); marked 'D' = tied rank 4 (12 species); C) Weighted Endemism; D) Phylogenetic Diversity; E) Phylogenetic Endemism; F) Imperiled

Phylogenetic Endemism (IUCN transformation). 679x759mm (96 x 96 DPI)

Page 27 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

. IUCN risk-weighted tree of lemurs (left), Isaac risk-weighted tree (right), and IUCN designation in the centre (note scale difference between trees).

348x245mm (150 x 150 DPI)

Page 28 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

The heatmaps of IPEIUCN100 (A) and IPEIsaac (B). The top 20 with the highest ranking are numbered. 451x380mm (96 x 96 DPI)

Page 29 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only0.00 0.05 0.10 0.15 0.20 0.25

0.00

0.05

0.10

0.15

A

Phylogenetic Endemism

IPE

IUC

N

0.00 0.05 0.10 0.15 0.20 0.25

0.00

0.04

0.08

B


IPE

Isaa

c

0.00 0.05 0.10 0.15

0.00

0.04

0.08

C

IPE IUCN100

IPE

Isaa

c

0 1 2 3 4 5 6 7

0.00

0.10

0.20

D

Phylogenetic Diversity

Phy

loge

netic

End

emis

m

0 1 2 3 4 5 6 7

0.00

0.04

0.08

E


IPE

Isaa

c

0 1 2 3 4 5 6 7

0.00

0.05

0.10

0.15

F


IPE

IUC

N

Page 30 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

●●●●

●

●

●

●

●●

●●

●

●

●

● ●

●

●

●

●

●●

●●

●

●

●

●

●

●

●

●●●

●

●●

●

●

●

● ●●

●●

●

●

●●

●

●●● ●

●

●●

●

● ●●

● ●●●

●● ●

●●●

● ●●

●●●

●

●●

●

●●●

●●

●●

●●

●

●

●●●

●

●●

●

●

●

●

●

●

●

●

●●

●

●

●

●●

●

●

●

●

●

●●

● ●

●

●

●

● ●●●●

●

●

●●

●●●●

●

●●

●●●●

●●●

●

●● ●

●

●●

●

●●

●

●●● ●●

●

●●

●

●●●● ●●

●

●

●●●●

●

●●

●●●●

●●

●●

●●●● ●

●

●●

●●●●●

●

5 10 15

0.00

0.10

0.20

A

Species Richness

Phy

loge

netic

End

emis

m

●●●●

●

●

●

●

●

●

●

●

●

●

●● ●

●

●●

●

●

●

●●

●●

●

●

●

●

●

●

●

●●

● ● ●

●

●

●●●

●●

●

●

●

●●

●●●

●●

●●

●

● ●●● ●●●●

● ●

●●●● ●●●

●●

● ●●

●●●●●● ●●

●

●

●

●

●● ● ●● ●

●

●

●

●●

●

●

●●●

●

●

●

●●

●

●

●

●

●

●●

● ●

●●

●

● ●●●●

●

●

● ●●●●●

●

●●

● ●●●●●●

●

● ● ●

●

●●●

●●

●●●● ●●

●

●●

●

●●●

●● ●

●

●

●●●●●

●●●●●● ● ●

●●

●●●● ● ●

●●

●●●● ●

●

5 10 15

0.0

1.0

2.0

B

Species Richness

End

emis

m

●●●●

●

●

●

●

●

●

●

●

●

●

●●●

●

●●

●

●

●

●●

●●

●

●

●

●

●

●

●

●●

● ● ●

●

●

●●●

●●

●

●

●

●●

●●●●

●

●●

●

●●●●●●●●

●●

●●●●●●●

●●

● ●●

●●●●●● ●●

●

●

●

●

●●● ●● ●

●

●

●

●●

●

●

●●●

●

●

●

●●

●

●

●

●

●

●●

●●

●●

●

●●●●●

●

●

●●●●●●

●

●●

● ●●●●●●

●

●●●

●

●●●

●●

●●●●●●

●

●●

●

●●●

●●●

●

●

●●●●●

●●●●●●●●

●●

●●●●●●

●●

●●●● ●

●

0.00 0.10 0.20

0.0

1.0

2.0

C


End

emis

m

●●●●

●

●

●●

●●

●●

●

●

●

● ●

●

●●

●

●●

●

●

●●

●

●

●

●

●

●●●

●● ●

●

●

●

● ●●

●●

●

●

●●

●

●●● ●

●

●●

●

● ●●

● ●●

●

●● ●

●●●

● ●●●●

●

●

●●

●

●●●

●●

●●

●●

●

●

●●●

●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●●

● ●

●

●

●

●●●●●

●

●

●●

●●●●

●

●●

●●●●

●●●

●

●● ●

●

●●

●

●●

●

●●● ●●

●

●●

●

●●●●

●● ●

●

●●●●

●●

●●●●● ●

●●●

●●

●● ● ● ●●●●●● ●

●

5 10 15

0.00

0.10

D

Species Richness

IPE

_IU

CN

100

●●●●

●

●

●

●

●●

●

●

●

●●

● ●

●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●●

● ●●

●

●

●

● ●●

●●

●

●

● ●●

● ●● ●

●

●●

●

● ●

●

●

●●●

●

●●

●●

●

●●

●●●

●●

●

●

●

●●●

●●

●●

●●

●

●

●●

●

●●

●

●●

●

●

●

●

●

●

●

●

●

●

●●●

●

●

●

● ●

●●

●●

●

●

●

●

●●●●

●

●

●

●

●●●

●

●

●●

●

●●●●

●●

●●

●●●

●●

●

●●

●

●●●●●

●

●●

●

●●●● ●

●●

●

●●●

●●

●●

●●●●●

● ●●

●●

●● ● ●●

●●●●● ●

●

5 10 15

0.00

0.04

0.08

E

Species Richness

IPE

_Isa

ac

●●●●

●

●

●●

●●

●●

●

●

●

●●

●

●●

●

●●

●

●

●●

●

●

●

●

●

● ●●

●●●

●

●

●

●●●

●●

●

●

●●

●

●●●●

●

●●

●

●●●

●●●●

●●●

●●●

●●●●●●

●

●●

●

●●●

●●

●●

●●

●

●

●●●

●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●●

●●

●

●

●

●●●●●

●

●

●●

●●●●

●

●●

●●●●

●●●

●

●●●

●

●●

●

●●

●

●●●●●

●

●●

●

●●●●

●● ●

●

●●●●

●●

●●●●●●

●●●

●●

●●●● ●●●●●●●

●

0.0 0.5 1.0 1.5 2.0

0.00

0.10

F

Endemism

IPE

_ IU

CN

100

●●●●

●

●

●

●

● ●

●

●

●

●●

●●

●

●

●

●

●●

●

●

●●

●

●

●

●

●

●

●●

● ●●

●

●

●

●●●

●●

●

●

● ●●

●●●●

●

●●

●

●●

●

●

●●●

●

●●

●●

●

●●●

●●

●●

●

●

●

●●●

●●

●●

●●

●

●

●●●

●●

●

●●

●

●

●

●

●

●

●

●

●

●

●●●

●

●

●

● ●

●●

●●

●

●

●

●

●●●●

●

●

●

●

●●●●

●

●●

●

●●●●

●●

●●

●●

●●●

●

●●

●

●●●●●

●

●●

●

●●●●●

●●

●

●●●

●●

●●

●●●●●

● ●●

●●

●●●●●

●●●●●●

●

0.0 0.5 1.0 1.5 2.0

0.00

0.04

0.08

G

Endemism

IPE

_Isa

ac

●●●●

●

●

●

●

●

●

●

●

●

●

●●●

●

●●

●

●

●

●●

●●

●

●

●

●

●

●

●

●●

● ● ●

●

●

●●●

●●

●

●

●

●●

●●●●

●

●●

●

●●●● ●●●●

● ●

●●●● ●●●

●●

● ●●

●●●●●● ●●

●

●

●

●

●● ● ●● ●

●

●

●

●●

●

●

●●●

●

●

●

●●

●

●

●

●

●

●●

● ●

●●

●

● ●●●●

●

●

● ●●●●●

●

●●

● ●●●●●●

●

●● ●

●

●●●

●●

●●●● ●●

●

●●

●

●●●

●● ●

●

●

●●●●●

●●●●●●●

●

●●

●●●●● ●

●●

●●●● ●

●

0 1 2 3 4 5 6 7

0.0

1.0

2.0

H


End

emis

m

●●●●

●

●●

●

●●

●

●

●

●

●

●●

●

●

●

●

●●●

●

●

●

●

●

●

●●

●

● ●

●●

●

●

●

●●●

● ●

●●

●

●●

●●●●●

●

●●

●

●●

●

●●●●

●

●●

●●●

●●

●●●

●●

●●

●

●●●

●●

●●

●●

●

●

●●

●

●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●●●

●

●●

●

●

●

●●

●

●

●

●

●

●●●●

●

●

●

●

●●●

●

●

●●

●

●●●

●

●●

●

●●

●●●●

●

●●

●

●●●

●●

●

●●● ●●●

●●

●●

● ●●●

●

●

●

●●●

●●

●

●

●

●

●●●●●

●

●

●●●●●

●

●

0 1 2 3 4 5 6 7

510

15I


Spe

cies

Ric

hnes

s

Page 31 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

Scatterplots of across-species measures of 'rarity' for Malagasy lemuriformes. 'Pendant Edge' is the relative length of the external branch leading from the species to the rest of the phylogenetic tree; Range size is the number of grid cells occupied by a species. P(ext) is the transformation of IUCN threat status from Table 1

of the main text. 265x257mm (150 x 150 DPI)

Page 32 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

Supplementary Information 1

Legends for the supplementary figures: 2

3 Figure S1. IUCN risk-weighted tree of lemurs (left), Isaac risk-weighted tree (right), and IUCN designation in 4 the centre (note scale difference between trees). 5 6 Figure S2: The heatmaps of IPEIUCN100 (A) and IPEIsaac (B). The top 20 with the highest ranking are numbered. 7 8 Figure S3: Scatterplots across the three phylogenetic conservation measures and Phylogenetic Diversity across 9 206 grid cells for Malagasy lemuriformes. A. IPEIUCN vs. PE; B. IPEIsaac vs. PE; C. IPEIsaac vs. IPEIUCN; D. PE vs. 10 PD; E. IPEIsaac vs. PD; F. IPEIUCN100 vs. PD. See main text for abbreviations. 11 12 Figure S4: Scatterplots of phylogenetic vs. nonphylogenetic conservation measures across 206 grid cells for 13 Malagasy lemuriformes. Endemism = Weighted Endemism; IPE = Imperiled Phylogenetic Endemism. See 14 Table 1 in main text for further details. 15 16 Figure S5: Scatterplots of across-species measures of 'rarity' for Malagasy lemuriformes. 'Pendant Edge' is the 17 relative length of the external branch leading from the species to the rest of the phylogenetic tree; Range size is 18 the number of grid cells occupied by a species. P(ext) is the transformation of IUCN threat status from Table 1 19 of the main text. 20 21

22

Page 33 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

Table S1. Taxonomic and threat status data for Malagasy Lemuriformes1.

Taxon

Sequence

data Sister to

Threat

status

DD

designatio

n Reason for DD designation

Allocebus trichotis yes DD EN habitat destruction

Avahi betsileo no

A.

peyrierasi DD EN possible habitat loss

Avahi cleesei yes EN EN

Avahi laniger yes LC LC

Avahi meriodionalis no

A.

peyrierasi DD NT restricted to a reserve

Avahi occidentalis yes EN EN

Avahi peyrierasi no A. laniger DD EN possible habitat loss

Avahi

ramanantsoavani no

A.

meridional

is DD NT restricted to small area

Avahi unicolor yes DD EN

slash-and-burn, charcoal production,

hunting

Cheirogaleus

adipicaudatus no C. medius DD EN possible habitat loss

Cheirogaleus crossleyi yes DD EN possible habitat loss

Cheirogaleus major yes LC LC

Cheirogaleus medius yes LC LC

Cheirogaleus

minusculus no C. major DD EN deforestation in main area

Cheirogaleus ravus no C. major DD EN deforestation in main area

Cheirogaleus sibreei yes DD EN deforestation in main area

Daubentonia

madagascariensis yes NT NT

Eulemur albifrons yes VU VU

Eulemur cinereiceps yes EN EN

Eulemur collaris yes VU VU

Eulemur coronatus yes VU VU

Eulemur fulvus yes NT NT

Eulemur macaco yes VU VU

Eulemur mongoz yes VU VU

Eulemur rubriventer yes VU VU

Eulemur rufifrons no E. rufus NT NT

Eulemur rufus yes DD EN slash-and-burn, logging and hunting

Eulemur sanfordi yes EN EN

Hapalemur aureus yes EN EN

Hapalemur griseus yes VU VU

Hapalemur

meridionalis yes VU VU

Hapalemur

occidentalis yes VU VU

Indri indri yes EN EN

Lemur catta yes NT NT

Page 34 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

Lepilemur

ankaranensis yes EN EN

Lepilemur dorsalis yes DD EN

rice and coffee cultivation, logging and

hunting

Lepilemur edwardsi yes VU VU

Lepilemur jamesorum yes DD NT restricted to small area

Lepilemur leucopus yes DD EN

burning, over-grazing and chargoal

production

Lepilemur

randrianasoli yes DD EN possible habitat loss

Lepilemur

sahamalazensis yes DD EN logging, charcoal production and hunting

Lepilemur

septentrionalis yes CR CR

Microcebus berthae yes EN EN

Microcebus

griseorufus yes LC LC

Microcebus murinus yes LC LC

Microcebus myoxinus yes DD EN possible habitat loss

Microcebus

ravelobensis yes EN EN

Microcebus rufus yes LC LC

Microcebus

sambiranensis yes EN EN

Microcebus tavaratra yes EN EN

Mirza coquereli yes NT NT

Mirza zaza yes DD EN

logging, slash-and-burn, but they do live in

forest fragments

Phaner electromontis no P. furcifer VU VU

Phaner furcifer yes LC LC

Phaner pallescens yes LC LC

Phaner parienti no P. furcifer VU VU

Prolemur simus yes CR CR

Prophithecus candidus yes CR CR

Prophithecus

coquereli yes EN EN

Propithecus coronatus yes EN EN

Propithecus deckeni yes VU VU

Propithecus edwardsi yes EN EN

Propithecus perrieri yes CR CR

Propithecus tattersalli yes EN EN

Propithecus verreauxi yes VU VU

Varecia rubra yes EN EN

Varecia variegata yes CR CR

1. The reason for DD designation is based on the habitat information provided by IUCN 2010

(www.iucnredlist.org, last accessed on 23-2-2011).

Page 35 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

Table S2. Genbank Accession numbers for Malagasy Lemuriforme sequence data.

Species original Adora3-3 Adora3-2 FGA ENO1 FIBA IRBP SWS

Allocebus trichotis

Avahi betsileo

Avahi cleesei

Avahi laniger

Avahi meriodionalis

Avahi occidentalis

Avahi peyrierasi

Avahi ramanantsoavani

Avahi unicolor

Cheirogaleus adipicaudatus

Cheirogaleus crossleyi GQ243528

Cheirogaleus major EU825408 AF271421

Cheirogaleus medius EU342218 EU342243 EU825388

Cheirogaleus minusculus

Cheirogaleus ravus

Cheirogaleus sibreei GQ243480 GQ243518

Daubentonia madagascariensis EU342219 EU342244

Eulemur albifrons

Eulemur cinereiceps

Eulemur collaris

Eulemur coronatus

Eulemur fulvus

Eulemur macaco EU342252

Eulemur mongoz EU342227 EU342253 AF081064

Eulemur rubriventer EU342228 EU342254 AF081065

Eulemur rufifrons

Eulemur rufus

Eulemur sanfordi

Hapalemur aureus

Hapalemur griseus EU342230 EU342255 AF081057

Hapalemur meridionalis

Hapalemur occidentalis

Indri indri

Lemur catta EU342231 EU342256

Lepilemur ankaranensis

Lepilemur dorsalis

Lepilemur edwardsi

Lepilemur jamesi

Lepilemur leucopus

Lepilemur randrianasoli

Lepilemur sahamalazensis

Lepilemur septentrionalis

Microcebus berthae GU231318 DQ003350 EF052288

Microcebus griseorufus GU231312 GU232114 GU231349

Microcebus murinus GU230994 EU342259 AF081054

Microcebus myoxinus GU231330 GU232130 GU231431

Microcebus ravelobensis GU231060 EU342260 EF052345

Microcebus rufus GU231068 DQ003458 EF052326

Microcebus sambiranensis GU231226 DQ003416 GU231511

Microcebus tavaratra GU231130 DQ003425 EF052355

Page 36 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

Mirza coquereli EU835930 EU342234 EU342261 EU835927 AY434080

Mirza zaza EU835914 EU835912

Nycticebus pygmaeus

Otolemur crassicaudatus

Phaner electromontis

Phaner furcifer

Phaner pallescens

Phaner parienti

Prolemur simus

Prophithecus candidus

Prophithecus coquereli

Propithecus coronatus

Propithecus deckeni

Propithecus edwardsi

Propithecus perrieri

Propithecus tattersalli EU342239

Propithecus verreauxi

Varecia rubra EU342241 EU342267 AF081055

Varecia variegata EU342242 EU342268

Species original VWF7 VWF11 VWF12 COII COIII Cytb 12S D-loop

Allocebus trichotis AF224620 AY44146 EU779975

Avahi betsileo

Avahi cleesei DQ856116 EF103295 DQ856036

Avahi laniger AY569208 AF224598 AY441453 AY254047 EF491652

Avahi meriodionalis

Avahi occidentalis AY584483 DQ856125 EF103294 AF474241 DQ856046

Avahi peyrierasi

Avahi ramanantsoavani

Avahi unicolor DQ856112 DQ856035

Cheirogaleus adipicaudatus

Cheirogaleus crossleyi EU825610 EU825510 AY605927

Cheirogaleus major EU825555 AY584487 AY582563 EU825344 AF474240 AY254050

Cheirogaleus medius EU825537 AY584486 AY582562 EU825334 AY192626 AY584498

Cheirogaleus minusculus

Cheirogaleus ravus

Cheirogaleus sibreei GQ243552 GQ243499

Daubentonia madagascariensis AY434049 L22776 AF224642 AM905039 AB371085

Eulemur albifrons AF081043 AF224568 AF081048 AF081034

Eulemur cinereiceps AF224562 AF175858 AF175799 EU333231

Eulemur collaris AF081041 AF224560 U53576 AF175775 EU333233

Eulemur coronatus AF224523 AY441448 AF175773 AF175861

Eulemur fulvus AF224534 AF175842 AF258003

Eulemur macaco L22777 AF224529 AF175790 EU333242

Eulemur mongoz AY434044 AF081045 EF552603 AF081051 AY043338 EU333182

Eulemur rubriventer AY434045 AF081046 AF224526 AF081052 AY043337 AF081038

Eulemur rufifrons

Eulemur rufus AF081042 AY582561 U53577 AF474234 AF081033

Eulemur sanfordi AF224561 AF175846 AF258016

Hapalemur aureus AY515558 AF224582 AY441446 AF474239 AY584489

Hapalemur griseus AY434047 AY569204 AF224572 AJ430044 AY582717 AY584491

Hapalemur meridionalis AY441447 AJ429206

Hapalemur occidentalis AY569205 AF224580 AJ428983 AY582718 AY584493

Page 37 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

Indri indri DQ856049 AY441455 AY043340 DQ855966

Lemur catta AY434046 L22780 AF224570 U53575 AF038013 AF175503

Lepilemur ankaranensis DQ529743 EF686696 DQ529451 DQ847455

Lepilemur dorsalis AY582592 EF686706 AY192625 DQ529594

Lepilemur edwardsi DQ529614 EU200437 DQ529325 DQ529473

Lepilemur jamesi EU200438

Lepilemur leucopus DQ529710 DQ109007 DQ529418 DQ529568

Lepilemur randrianasoli DQ529661 DQ451105 DQ529372 DQ529520

Lepilemur sahamalazensis DQ529738 DQ529458 DQ529446 EF686752

Lepilemur septentrionalis DQ529731 DQ109021 DQ529439 DQ529589

Microcebus berthae EF052413 GU326978 EF175238 GU327166 AF285466

Microcebus griseorufus EF052421 GU327148 AY582661 AY167076 AY582701 AF285490

Microcebus murinus GU326994 AF224628 AF285565 AY582696 EF175283

Microcebus myoxinus EF052431 GU327018 GU327205 AF285459

Microcebus ravelobensis EF052466 AY434034 GU327031 AY582635 GU327219 AY582712 AY159695

Microcebus rufus EF052455 GU327048 AY582653 GU327231 AY582713 AY159722

Microcebus sambiranensis EF052470 AY569191 AY582636 GU327251 AY582676 AY159704

Microcebus tavaratra EF052479 GU327057 DQ534993 AF285533 AF285456

Mirza coquereli EU835934 EU835931 AF224623 DQ093175 AJ429628 EU779978

Mirza zaza EU835918 EU835915 EU779961 DQ093169 EU779977

Nycticebus pygmaeus

Otolemur crassicaudatus

Phaner electromontis

Phaner furcifer AY441456

Phaner pallescens EU779960 EU779976

Phaner parienti

Prolemur simus AY515559 AY582548 AJ428977 AF474238 AY584488

Prophithecus candidus AF356204

Prophithecus coquereli AF285492 AF224613 AF285528 DQ855971

Propithecus coronatus AF224610 AY441451

Propithecus deckeni AF224606 DQ855973

Propithecus edwardsi AY584484 AY582556 AF474236 AF354722

Propithecus perrieri DQ855968

Propithecus tattersalli AY434038 L22782 AF224600 U53573 AF175793 DQ855972

Propithecus verreauxi AY584485 AY582557 AJ429620 AF354711

Varecia rubra AY434048 L22785 AF224592 AY441450 AF175791 AF081028

Varecia variegata AF081040 AF224587 AF081047 AF474235 AY584494

Page 38 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only

Table S3. Comparisons between variables using linear and nonlinear (quadratic) models. Abbreviations as

in the main text.

Model Linear R2 F DF P-Value

Non-

linear R2 F DF P-Value

PE~SR 0.6738 424.5 204 <0.001 0.7004 240.6 203 <0.001

WE~SR 0.4049 140.5 204 <0.001 0.4593 88.06 203 <0.001

PE~WE 0.7937 789.6 204 <0.001 0.8393 536.3 203 <0.001

IPE_IUCN~SR 0.6504 382.4 204 <0.001 0.682 220.8 203 <0.001

IPE_Isaac~SR 0.7808 731.3 204 <0.001 0.7884 383 203 <0.001

IPE_IUCN~WE 0.6692 415.7 204 <0.001 0.7064 247.6 203 <0.001

IPE)Isaac~WE 0.6452 373.8 204 <0.001 0.7099 251.9 203 <0.001

Page 39 of 39



123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Review Only - Simon Fraser Universityamooers/papers/Gudde_etal_DD.pdf · 2012. 7. 21. · For...

Documents

Transcript of For Review Only - Simon Fraser Universityamooers/papers/Gudde_etal_DD.pdf · 2012. 7. 21. · For...