Macroecological correlates of Darwinian

shortfalls across terrestrial vertebrates

Jhonny J. M. Guedes1, José Alexandre F. Diniz-Filho2 and Mario R. Moura3,4
1
Departamento de Ecologia, Programa de Pós-Graduação em Ecologia e Evolução, Universidade Federal de
Goiás—Campus Samambaia, Goiânia, GO 74690-900, Brazil
2
Departamento de Ecologia, Instituto de Ciências Biológicas, Universidade Federal de Goiás — Campus
Research Samambaia, Goiânia, GO 74690-900, Brazil
3
Departamento de Biologia Animal, Universidade Federal de Campinas, Campinas, SP 13083-970, Brazil
4
Departamento de Biociências, Universidade Federal da Paraíba, Areia, PB 58397-000, Brazil
Cite this article: Guedes JJM, Diniz-Filho JAF,
Moura MR. 2024 Macroecological correlates of JJMG, 0000-0003-0485-3994; JAFD-F, 0000-0002-0967-9684; MRM, 0000-0002-7369-7502

Darwinian shortfalls across terrestrial vertebrates.

Most described species have not been explicitly included in phylogenetic
Biol. Lett. 20: 20240216.
trees—a problem named the Darwinian shortfall—owing to a lack
https://doi.org/10.1098/rsbl.2024.0216 of molecular and/or morphological data, thus hampering the explicit
incorporation of evolution into large-scale biodiversity analyses. We
investigate potential drivers of the Darwinian shortfall in tetrapods, a
Received: 19 April 2024 group in which at least one-third of described species still lack phylogenetic
Accepted: 18 June 2024 data, thus necessitating the imputation of their evolutionary relationships
in fully sampled phylogenies. We show that the number of preserved
specimens in scientific collections is the main driver of phylogenetic
knowledge accumulation, highlighting the major role of biological
Subject Category: collections in unveiling novel biodiversity data and the importance of
Evolutionary biology continued sampling efforts to reduce knowledge gaps. Additionally,
large-bodied and wide-ranged species, as well as terrestrial and aquatic
amphibians and reptiles, are phylogenetically better known. Future efforts
Subject Areas: should prioritize phylogenetic research on organisms that are narrow-
ecology, evolution ranged, small-bodied and underrepresented in scientific collections, such as
fossorial species. Addressing the Darwinian shortfall will be imperative for
advancing our understanding of evolutionary drivers shaping biodiversity
Keywords: patterns and implementing comprehensive conservation strategies.
biodiversity shortfall, evolutionary knowledge,
genetic data, phylogenetic relationship, tetrapod

1. Introduction
Author for correspondence:
In a rapidly changing world where biodiversity is being lost at unparal-
Jhonny J. M. Guedes leled rates [1], biodiversity knowledge shortfalls [2] can hinder the effec-
e-mail: [email protected] tive implementation of data-driven conservation strategies [3]. For instance,
only a small fraction (about 20%) of all extant species on Earth have been
named [4,5], and for most described species the available information is
incomplete and biased towards charismatic or easily accessible organisms [6–
8]. These knowledge shortfalls, either individually or through their interac-
tions [3,9], beset our ability to better understand biodiversity patterns and
their underlying processes. For example, mechanisms like mutation, gene
flow, natural selection and genetic drift have shaped the genetic inheritance
that species share. Hence, comparative analyses must consider the shared
evolutionary history among taxa since species are not independent of each
other [10]. But for most species, their placement in the tree of life is largely
unknown—a problem named the Darwinian shortfall [11]—especially owing
to a lack of comprehensive molecular and/or morphological data [12]. This
biodiversity shortfall hampers the explicit incorporation of evolution in
large-scale conservation and biodiversity analyses, which could lead to biased
results and impair conservation practices [11,13,14].
Electronic supplementary material is available
online at https://doi.org/10.6084/
m9.figshare.c.7358195.

© 2024 The Author(s). Published by the Royal Society. All rights reserved.
 The Darwinian shortfall can be divided into three main components: (i) the lack of data on phylogenies, that is on basic 2
relationships among species, (ii) high uncertainties in divergence time estimates owing to a lack of knowledge about branch
lengths and temporal calibration of diversification events, and (iii) a paucity of models of trait evolution [11]. The latter two

royalsocietypublishing.org/journal/rsbl Biol. Lett. 20: 20240216

components are conditioned to the former, and therefore, looking at phylogenetically neglected taxa represents an interesting
avenue to better understand phylogenetic knowledge accumulation and reduce data gaps [13,15]. Although complete species-
level phylogenies exist for major groups of organisms, such as tetrapods [16–20], many species still lack basic phylogenetic
data and are taxonomically imputed to create ‘fully-sampled’ phylogenies. Gaining insights into the patterns and underlying
determinants of the Darwinian shortfall can offer guidelines to mitigate significant knowledge gaps.
Tetrapods are one of the most well-studied organisms globally [7,21], which improves species-level trait data availability
compared to neglected taxa and allows macroecological investigations to be conducted with relatively good taxonomic
coverage. However, a recent study showed that even for these well-known organisms, nearly one-quarter of more than 34
000 species do not have a single genetic sequence available in widely used genetic repositories [15]. The Darwinian shortfall
may be even more pronounced since genetic and genomic data alone do not equate to knowledge of species placement
within phylogenies [22]. Indeed, on average 38% of tetrapod species in fully sampled phylogenies have their phylogenetic
placement imputed (30.7% of mammals up to 43.9% of amphibians and squamates [18–20]). Understanding how species’
traits and geography make them more likely to be phylogenetically neglected could help identify organisms or regions that
have been phylogenetically overlooked and optimize future research efforts. Here we investigate how biological, geographic,
socioeconomic and conservation-related factors (see box 1) may affect the probability of a species being imputed in phylogenetic
reconstructions.
Identifying which tetrapod species are more likely to be phylogenetically overlooked can help guide the development of
new initiatives to address the Darwinian shortfall, ultimately benefiting comparative biological research. The vast number of
museum specimens housed in scientific collections worldwide [32], despite many not being stored in DNA-friendly substances,
now present a valuable resource due to the rapid development of tools and techniques for analysing genetic data from these
specimens [46,47]. By revealing attributes of phylogenetically neglected groups, we open an interesting avenue to resolving the
tree of life, one of the most complex challenges in all biology.

2. Methods
(a) Phylogenetic assessment and species-level covariates
We first identified which tetrapod species had their evolutionary relationships imputed in fully sampled phylogenies available
for the group [16–20], thus creating a binary response variable for 33 281 species included in such phylogenies. We excluded 228
marine species (according to the International Union for Conservation of Nature (IUCN)), keeping only terrestrial vertebrates
(n = 33 053) in subsequent analyses. We then selected eight putative predictors to investigate how species’ biology, geography,
biodiversity appeal and socioeconomic-related factors affect the probability of species being imputed on phylogenies. Many
predictors used here come from the TetrapodTraits, a recent database for the world’s tetrapod including standardized spe-
cies-level attributes [48]. Here, we outline the attributes used as predictors, along with brief explanations of how they were
computed.
The year of species description was based on the original description publication date and was available for all but one
species (Indotyphlops pushpakumara, an undescribed species included in Tonini’s phylogeny [19]). For body size, we used body
mass information for birds, mammals and reptiles as data coverage was on average higher than 95% (n = 24 761 out of 25 815
species). Conversely, body mass information was available for only 21.4% (n = 1547) amphibian species, and therefore, we used
body length, which represented 97.9% (n = 7083) species. Information on microhabitat use was available for 31 501 (95.3%)
species and it was converted into a continuous metric of verticality [49], with species being scored as 0 = strictly fossorial, 0.25
= fossorial and aquatic/terrestrial or fossorial and aquatic and terrestrial; 0.5 = aquatic/terrestrial, or fossorial and arboreal or
fossorial and aquatic/terrestrial and arboreal; 0.75 = terrestrial/aquatic and arboreal, or terrestrial and aquatic and arboreal or
terrestrial and aerial and 1 = strictly arboreal or aerial. Species-specific sources on body size and microhabitat are available in
[48].
Some TetrapodTraits attributes were derived as within-range predictors based on expert-based range maps for amphibians
[50–52], birds [17,52], mammals [51–53] and reptiles [16,51,54] overlaid onto a 110 × 110 km cylindrical equal-area grid cell
scheme. This grain size minimizes false presences related to the use of expert range maps [55]. Species range size was represen-
ted as the number of 110 × 110 km grid cells overlapped by range maps. Two other attributes corresponded to within-range
averages of raster layers aggregated to the resolution of the grid cell scheme, namely elevation, which was based on a 1 km
topographic layer [56] and roughly captures broad elevational patterns in a continuous format, without the discretization of
species into lowland versus highland areas and human density (as inhabitants per km2 in year 2017), derived from the HYDE
3.2 database at a spatial resolution of 5 arc-min [57]. Additionally, we obtained endemism richness, which represents a proxy for
range rarity. This metric was computed as the sum of the inverse range sizes of all species per taxonomic class in a grid cell [58],
which we then used to calculate the median endemism richness for each species based on the cells they occur in.
We extracted the number of preserved specimens per species deposited in biological collections worldwide using the
function occ_count in the rgbif R package [59]. For that, we created search queries containing valid species name plus their
unique synonyms (i.e. invalid names that can be traced back to a single valid name) and set the basisOfRecord argument to the
preserved specimen. Synonym information was obtained from the IUCN taxonomy backbone using the rl_synonyms function in
 3
Box 1. Potential drivers that may affect the probability of a species being imputed in available phylogenies.

royalsocietypublishing.org/journal/rsbl Biol. Lett. 20: 20240216

Species-level variation in the probability of being directly accounted for (or neglected) in phylogenetic reconstructions may
be driven by biological, geographical, conservation and socioeconomic-related factors. In this box, we explain the rationale
for each hypothesis and their expected effects.

Species biology
Biological attributes can affect species’ detectability and researchability, as well as sampling logistics, which can increase the
chances of a species being explicitly included in phylogenies.
— Body size. Larger animals are easier to detect, collect and study [23–25]. They also tend to attract more scientific,
societal and governmental attention [26–28]. We expect that larger animals will be less likely to be phylogenetically
imputed than small-bodied organisms.
— Verticality. Terrestrial and aquatic species are more easily found and collected in their microhabitats than fossorial,
arboreal or aerial species using standard sampling techniques [29,30]. For instance, fossorial reptiles are usually
underrepresented in scientific collections [31], which influences the relatively low research effort directed towards
this group [27]. As limited sample sizes may limit the chances of a species being phylogenetically evaluated, we
anticipate an inverse hump-shaped relationship between verticality and the probability of a species being imputed on
phylogenies, i.e. higher likelihood for species positioned at the extremes of the verticality spectrum.
— Number of preserved specimens. Several millions of preserved specimens are currently housed in scientific collections
[32,33], with the number of preserved specimens potentially mirroring species abundance in nature [34]. These
specimens form the basis for most subsequent scientific investigations [35,36]. We hypothesize that species better
represented in scientific collections will be less likely to be phylogenetically imputed.

Geography and biodiversity appeal

Species’ range size and the spatial patterns in species distributions (i.e. regions of high conservation appeal or areas of low
accessibility) can influence the likelihood of an organism being found and studied [8,37].
— Range size. Wide-ranged species have more chances of being detected and collected across their ranges [25,38], besides
being typically more abundant than narrow-ranged species [39], thus making them more accessible for molecular
data and phylogenetic studies [15]. We hypothesize that wide-ranged species will be phylogenetically better known
than geographically restricted species.
— Elevation. Mountainous regions have limited accessibility relative to lowland areas, where the latter have often been
intensively explored first and thus their biota are better known [8,27]. We expect that highland species will show
higher chances of being phylogenetically imputed than species occurring at lower elevations.
— Endemism richness. Scientists may prefer to work in areas of high conservation appeal, i.e. with many rare or range-
restricted species [8,40]. However, these areas can have many species with low detectability and, sometimes, legal
restrictions prohibiting or limiting specimen collection [15]. Therefore, we anticipate that the probability of being
phylogenetically neglected is positively related to endemism richness.

Socioeconomics
Socioeconomic factors can influence phylogenetic research through the availability of either researchers or the infrastruc-
ture needed for conducting studies, as well as through effects on sampling logistics and financial constraints.
— Human density. Species occurring in highly populated areas may have increased detectability as these areas are more
accessible to researchers, who usually work under many logistics and financial constraints, particularly in developing
countries [41]. Hence, remote and usually less populated areas—and the species therein—are visited less frequently
by researchers [42,43]. We expect that species living close to densely populated areas will have a lower probability of
being phylogenetically imputed.
— Year of description. Early described species usually receive more scientific attention than either recently described taxa
or those yet to be named [8,27,44]. Therefore, time is clearly needed for accumulating scientific knowledge about the
described diversity [45], contributing to their proper inclusion in the tree of life. However, the process of collecting
tissue samples is common in recent collections. In contrast, older specimens, especially from rare species, often lack
suitable genetic material owing to historical and methodological constraints or storage in media that no longer permit
or hinder extraction. Thus, the year of description can influence the accumulation of phylogenetic knowledge in
contrasting ways.

the rredlist R package [60]. Data on these—and the spatially based—predictors were available for more than 99.9% of species.
After excluding species with missing data in at least one predictor, our final dataset included 30 321 species: 6526 amphibians,
9369 birds, 5052 mammals and 9374 reptiles.

(b) Statistical analyses

We modelled the probability of a species being phylogenetically imputed (binary response variable: 0 = not imputed, 1 =
imputed) as a function of eight putative predictors (fixed effects; see box 1) while accounting for taxonomic relatedness (random
effects) using generalized linear mixed-effects models (GLMM; [61]) with a binomial error distribution and a logit link function
 (a) (b) 4

Proportion of phylogenetically

22 0

45
1
2

5
2

323

5
3
1

4
9
1
6

1
6
3
8
1
1

4
3

royalsocietypublishing.org/journal/rsbl Biol. Lett. 20: 20240216

39

53
77
13
42
37

24
60
11
48
15
11
52
12
19
44
12

15
13

16
16

11
21
49
16

97

17
11
21
41
16

23

12
27
20
28

32
14
33
28
42
12
86
14
1.0 1.0

53
47

30
10
19
9

4
4

9
4

7
imputed species
0.5 0.5

0.0 0.0
ra hy sot e
Ty batr hii ae

Cy H nec dae
Ce ora odidae
ut mp ae

C an e
e
Rh lma opi e
in tob dae
yc m ae
Cr tra eni ae
St uga hyli ae
O Cer om rid e

e
A ype ynid e
ro liid e
Le icr epti ae
o y e
H erm tyl ae
eo ph ae
eg yn ae
M B hry ae
Rh batr nid e
ac ac ae
ho ae
Ra Hyl ae
xa e
ae
ph ac dae

do at an ae

en pi ae
Cy Mic sau dae
Le dro apid e
io ph ae
A U pha dae
T isb lti e
Le rop aen dae
ot p ae
X An lop ae
od ie ae

A atid e
e
om am e
Co ops ae
no ph brid e
Tr phth opi ae
on m e
hi e
ha E eii ae
La oda pid e
m cty ae
e
bl ro ae
ha ae
Le Sci idae

ol ri e
Ph A em ae
lo m e
ct ae
ae
Ce lcht mi iida

m da
Si hik tida
a

ra st da

op hy a
H hyr rida
rth ro a

pt oh da

yo uf ida

ni ida

lin rel rida

ph pe da

m a
ng a
H Dib uida

ym Ty lu da

og al da
op ida

er la da

Po rop lida

Li sau ida

yl ga ida
to op id

cl yl tid

Te hon ilid

Pt tre iid
Ba had atid
a c d

nt op tid

D dac lid
el o id
M phr iid
op id

op hid
rid

lid

X rho did

pt ido id
h id
en n id
er llid

al id

T id

Eu ych hiid
ep tid

a d

da id
id
lo hi

ho hi

o li

ce ii
m ro l i

i
yp hi

r
io nc

yl
H ecil

b o

er i
G alep

l
M l
Ca

p
e

m
a

l
no
A

Sp
G
(c) (d)
Proportion of phylogenetically

2
5
3

26

16

13
44

32
21
94
67
25

10
36
58
19
11
11
50
14
17
23
37
39

28
35
17
17

32
22
10

10
11
33
43

13
98
51
88
11
15
14

45
11
82
27
62
10

57
1.0 1.0

7
8

7
5
5

3
4

9
9

4
imputed species

0.5 0.5

0.0 0.0
Ci Ind lbu dae
O loso tor ae
nt at ae
ho ae
lo ng ae
eg se e
Ci in ae
co ae
ro al ae
ra ae
oe am e
Re culi ae
i e
V loce ae

N Ap nidae
ec od e
er ii e
Co T clid ae
po ic ae
a e
Pa peti e
Cr sser ae
A ticid e
C luad ae
rm b e
M cari ae
il ot e
M epi idae
ill e
ae

lo uri e
N ysci e
D roc rida e

Ta Pit octi ae
ch he dae
Te ossi ae
as ec e
M pod dae
Ph M ssid e
al ur ae
Ct T eri ae
od sii e
Pr acty dae
Tr avii ae
S ul e
N ori idae

T y e
pr aii e
H yi e
D tri e
t d e
Le erop rgid e
V pile odi ae
pe u e
M Ca tili idae
ec ch ae
ha ae
te e
ae
as om e
A rop ida

Ph Tin cida

m da

o a

Pt arin ida

ph ida
a

ac ida

Fo olum ida

Ph om ida

ac a

Ca mal rida
m da
A yc da

D nr da

o a

en ar da

ag da

om da
Ca up ida
om da
a
Ba ipo cida
Pt hye ida

es m da

A ida
nc ica lid
do m id
op id
Ch Va rid

ith id
sti id
ac R lid
co lid

ni id

P zid
ire id

o d
no urn id

Eu gid
d

i id

ot ttid
id

r d

l d

ol id

an id

ys d

m tri id
op id

lid
a i

yp i

yg cii

y i

oc li

es ci
G con

r r
yr lli on

g
i

b e
no Gl

g
c

t
Bu

A
al
Ph

Figure 1. Top 30 families with knowledge gaps on species evolutionary relationships. Per family proportion of phylogenetically imputed species across (a) amphibians,
(b) reptiles, (c) birds and (d) mammals. The number above the bars indicates total species richness per family. Silhouette illustrations reproduced from phylopic.org.

[62]. We included the taxonomic family as a random variable in our models to minimize dependence issues among species.
Families with less than three species (electronic supplementary material, table S1) were removed to reduce instability in model
estimates [63], which decreased the number of species to be modelled to 30 178.
We evaluated whether phylogenetic regression models were needed by examining the phylogenetic autocorrelation of
GLMM residuals through Moran’s I correlograms computed across 14 distance classes [10]. The phylogenetic correlograms were
based on averaged results from 50 fully sampled phylogenies for each taxonomic class [16–20]. For reptiles, we first constructed
supertrees by combining Tonini’s and Colston’s phylogenies [16,19] using the function tree.merger in the R package RRphylo [64],
which preserves branch length information in the combined trees [65]. For the global models, we used only five trees owing
to computational limitations. Analyses of phylogenetic autocorrelation were performed using the R packages phylobase [66] and
phylosignal [67].
Prior to constructing the GLMMs, continuous predictors were log10 transformed if skewness or kurtosis were outside the
range of −2 and +2 [68], then centred and scaled (z-transformed) to allow direct comparisons of their effect sizes. We checked
for multicollinearity among predictors using the variation inflation factors (VIF), where strong multicollinearity is usually
attributed to VIFs > 5, indicating that variables should be removed from the analysis [69]. Since none of our continuous variables
had VIF > 4, we kept them all in subsequent analyses (electronic supplementary material, table S2).
We modelled our binary response variable separately for amphibians, birds, mammals and reptiles. Models were constructed
globally as well as separately for each biogeographic realm [70], except for Oceania owing to its low sample size (electronic
supplementary material, table S3). Species whose range overlapped realms by >70% were assigned to realm-scale models. We
inspected the model fit using the R package DHARMa [71] and assessed the explained variation by calculating the pseudo-R2
with the R package performance [72]. We used the package lme4 [73] for fitting the mixed-effect models and usdm [74] for
computing VIF values. All analyses were performed using R version 4.3.1 [75]. See data accessibility for raw data and R-code.
Finally, we computed a measure of ‘Darwinian deficit’ [76] per family and class to quantify the relative contribution of
phylogenetically imputed relationships in representing the accumulated evolutionary history across taxa. This measure is
based on Faith’s phylogenetic diversity (PD) metric [77] and was calculated as PD imputed species/(PD imputed species + PD
non-imputed species). The Darwinian deficit ranges from 0 to 1 and informs the proportion of total PD (i.e. the sum of branch
lengths) that is attributed to imputed species in a sample (e.g. family). We obtained the Darwinian deficit per family across 100
fully sampled phylogenies for each taxonomic class, using only families with at least one imputed species. We then inspected
whether average values per family were influenced by the respective species richness.

3. Results
The proportion of phylogenetically imputed species was highly variable across taxa (figure 1 and electronic supplementary
material, table S4). Some suborders (sensu reference [25]), such as Crocodylia, Perissodactyla and Sphenisciformes already have
all of their described species explicitly incorporated into the most recent fully sampled phylogenies available. Conversely,
coverage is still relatively low for many taxa, such as Gymnophiona (n = 127 imputed species out of 190, or 66%), Tinamiformes
(n = 27 out of 43, or 62.8%), Dibamoidea (n = 13 out of 21, or 61.9%) and Monotremata (n = 2 out of 4, or 50%). Across
 5
Amphibians
200
Reptiles

royalsocietypublishing.org/journal/rsbl Biol. Lett. 20: 20240216

Birds
Mammals
150

Density
100

50

0
0.36 0.39 0.42
Darwinian deficit
Figure 2. Density plots show the distribution of ‘Darwinian deficits’ computed across 100 phylogenetic trees for each taxonomic class. This metric ranges from 0 to 1
and informs the proportion of total phylogenetic diversity that is attributed to imputed species. Vertical dashed lines show mean values.

Global Afrotropic Australasia IndoMalay Nearctic Neotropic Palaearctic

Body size

Verticality2

N. preserved
specimens

Range size

Elevation

Endemism
richness

Human
density

Year of
description
–1 0 1 –2 –1 0 –4 0 4 –1 0 –2 –1 0 1
.0
.5

.5
.0
.5
0
5
0.

0.
0.
–1
–0

–1
–1
–0

Model Coefficients (CI 95%)

Figure 3. Influence of potential drivers on the chances of species being phylogenetically imputed. Standardized coefficients from the generalized linear mixed-effect
models fit to all data (global) and for each biogeographic realm are shown. Horizontal lines denote the 95% confidence intervals around each coefficient. Predictors
with significant effect sizes (p‐value < 0.05) are those whose horizontal lines do not cross 0. Coefficients missing in the Nearctic realm represent extreme and
non-significant values that were removed to improve visualization. See electronic supplementary material, table S6 for detailed information on model estimates.
Silhouette illustrations reproduced from phylopic.org.

tetrapod classes, the proportion of phylogenetically imputed species per family was unrelated to species richness (electronic
supplementary material, figure S1). Furthermore, imputed species contributed to approximately 35% of the total phylogenetic
diversity for birds and mammals, 41% for reptiles and 43% for amphibians (figure 2). At the family level, the Darwinian deficit
was not influenced by species richness (electronic supplementary material, table S5 and figure S2).
Global models explained between 47.6 and 53.4% of the variation in the probability of species being imputed in fully
sampled phylogenies, but this number varied from 43% to 93.4% in realm-specific analyses (electronic supplementary material,
table S6). GLMMs were mostly robust in terms of fit, and their residuals did not show phylogenetic autocorrelation (electronic
supplementary material, figures S3–S7). Across all tetrapod classes, the number of preserved specimens emerged as the most
important predictor, which had a negative effect on the chances of a species being phylogenetically imputed (figure 3 and
electronic supplementary material, figure S8). Large-bodied and wide-ranged species showed lower chances of being phyloge-
netically imputed across most class-realm combinations. Verticality (in the hump-shaped form) and year of description had a
positive relationship with our response variable globally, while the effects of elevation, human density and endemism richness
affected different taxonomic classes in contrasting ways (figure 3). The effect sizes of other predictors were also important, but
their direction and magnitude varied across tetrapod classes and realms (figure 3).
4. Discussion 6

Despite recent advances in the field of phylogenetics, major gaps and biases still remain in the tree of life [13,15], with about

royalsocietypublishing.org/journal/rsbl Biol. Lett. 20: 20240216

one-third of tetrapod evolutionary history being represented through imputations (figure 2 and electronic supplementary
material, figure S2). Imputed species often lack empirical data on morphology and genetics, thus having their phylogenetic
positions inferred based on various assumptions that may lead to potential inaccuracies and uncertainties in phylogenetic
reconstructions. Given that evolutionary dynamics influence biodiversity patterns at various scales, it is important to use
comparative methods to evaluate these patterns and investigate the evolutionary imprints on them [78–80]. However, imputed
phylogenetic relationships may introduce biases and errors in comparative biodiversity analyses and conservation planning
reliant on evolutionary information (e.g. [81]). While imputation techniques offer valuable insights into understudied taxa,
cautious interpretation and validation procedures are necessary to mitigate the potential negative impacts on our conclusions.
By investigating the macroecological determinants of the Darwinian shortfall across terrestrial vertebrates, we showed that
the availability of preserved specimens in scientific collections is by far the most important driver of phylogenetic knowledge
accumulation (figure 3). This finding highlights the major role—although yet underappreciated potential—of natural history
collections in unveiling novel biodiversity data [32]. Assuming that the number of preserved specimens mirrors species’ natural
abundances [34,82], it highlights the importance of continued sampling efforts—particularly towards understudied organisms
and under-sampled areas [8,28,83]—to bridge biodiversity knowledge gaps [84]. Furthermore, it shows that scientists should
direct research efforts towards underrepresented organisms deposited in scientific collections. In today’s scientific landscape,
molecular data have central roles in phylogenetic reconstructions [22], where boosted by recent advances in molecular analyses
and related tools, even old museum specimens can be genetically sequenced [46]. Moreover, tissues are now frequently frozen
for future genetic analyses when collecting specimens, which are, in turn, stored in DNA-friendly substances—opposite to
what was usually done until a few decades ago [36]. Therefore, current trends in the molecular arena may enable the explicit
inclusion of species with limited sample sizes in the tree of life.
Large-bodied and widespread species are less likely to be phylogenetically imputed (figure 3). These animals are more
easily detected during field work [8,25,38] and also attract more public and scientific interest compared to less-conspicuous
organisms [7,26]. Therefore, they are collected more often across their ranges, resulting in increased availability of morpholog-
ical and genetic data and facilitating their explicit inclusion in phylogenies. Moreover, we found that terrestrial and aquatic
herptiles were phylogenetically better known than fossorial and arboreal species (figure 3). Detection probability and collection
rates of species can be affected by their biology as well as the availability of sampling techniques [85–87], thus affecting
species researchability differently [27]. Combined, these findings suggest that future sampling efforts for collecting and
genetic assessment should focus on narrow-endemics and small-bodied preserved specimens housed in biological collections,
particularly those of fossorial, arboreal and aerial habitat use. Given that most species remain undescribed—many of which
have already been collected in nature [35,36,88]—directing efforts towards such neglected organisms may also help to uncover
hidden diversity and boost species descriptions [89–91].
At least one-quarter of known terrestrial vertebrates still lack genetic data [15], and if only considering phylogenetically
informative sequence data (e.g. [13]), this figure could be even higher. Among imputed species in recent fully-sampled tetrapod
phylogenies [16–20], we found that 45% (n = 4825 species) have genetic sequences available in public repositories, although
probably not informative enough for phylogenetic reconstructions. Furthermore, hotspots of genetically unsampled species [15]
mirror the spatial patterns of the Linnean shortfall (i.e. lack of knowledge about species’ taxonomy; [92]), both being more
pronounced in the tropics [25,90]. Thus, the proportion of genetically assessed species is probably overestimated, especially
in species-rich areas. Current biases and gaps in tetrapod phylogenetic knowledge could impact the outcomes of downstream
biodiversity analyses and conservation initiatives that incorporate phylogenetic trees [3,93,94]. Reducing the Darwinian shortfall
is crucial not only for improving our understanding of the evolutionary drivers of global biodiversity patterns, but also for
advancing the implementation of more comprehensive conservation strategies.

Ethics. This work did not require ethical approval from a human subject or animal welfare committee.
Data accessibility. Raw data and R-code to replicate the findings of this study are available at Dryad Digital Repository [95].
Supplementary material is available online [96].
Declaration of AI use. We have not used AI-assisted technologies in creating this article.
Authors’ contributions. J.J.M.G.: conceptualization, formal analysis, investigation, methodology, writing—original draft, writing—review and editing;
J.A.F.D.-F.: supervision, writing—review and editing; M.R.M.: supervision, writing—review and editing, visualization, resources.
All authors gave final approval for publication and agreed to be held accountable for the work performed therein.
Conflict of interest declaration. We declare we have no competing interests.
Funding. This work was supported by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) (finance code 001), as a
PhD scholarship provided to J.J.M.G. (proc. 88887.478942/2020-00); CNPq productivity fellowship and the National Institutes for Science and
Technology (INCT) in Ecology, Evolution and Biodiversity Conservation (grant numbers 465610/2014-5); Goiás Research Foundation (FAPEG)
(grant no. 201810267000023); and São Paulo Research Foundation (FAPESP) (grant nos. 2021/11840-6 and 2022/12231-6).
Acknowledgements. The authors are grateful to Gabriel Nakamura, Juliana Stropp, Luis M. Bini and Leandro Duarte for their comments on an
earlier version of this manuscript. J.J.M.G. thanks the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) for a
PhD scholarship. Work by J.A.F.D.-F. is supported by Goiás Research Foundation (FAPEG) and CNPq productivity fellowships, and has been
developed associated with the National Institutes for Science and Technology (INCT) in Ecology, Evolution and Biodiversity Conservation.
M.R.M. acknowledges support from São Paulo Research Foundation (FAPESP).
References 7

royalsocietypublishing.org/journal/rsbl Biol. Lett. 20: 20240216

1. Barnosky AD et al. 2011 Has the Earth’s sixth mass extinction already arrived? Nature New Biol. 471, 51–57. (doi:10.1038/nature09678)
2. Hortal J, de Bello F, Diniz-Filho JAF, Lewinsohn TM, Lobo JM, Ladle RJ. 2015 Seven shortfalls that beset large-scale knowledge of biodiversity. Annu. Rev. Ecol. Evol. Syst. 46, 523–
549. (doi:10.1146/annurev-ecolsys-112414-054400)
3. Diniz Filho JAF et al. 2023 Macroecological links between the Linnean, Wallacean, and Darwinian shortfalls. Front. Biogeogr. 15, 1–11. (doi:10.21425/F5FBG59566)
4. Caley MJ, Fisher R, Mengersen K. 2014 Global species richness estimates have not converged. Trends Ecol. Evol. (Amst.) 29, 187–188. (doi:10.1016/j.tree.2014.02.002)
5. Mora C, Tittensor DP, Adl S, Simpson AGB, Worm B. 2011 How many species are there on earth and in the ocean? PLoS Biol. 9, e1001127. (doi:10.1371/journal.pbio.1001127)
6. Etard A, Morrill S, Newbold T. 2020 Global gaps in trait data for terrestrial vertebrates. Glob. Ecol. Biogeogr. 29, 2143–2158. (doi:10.1111/geb.13184)
7. Troudet J, Grandcolas P, Blin A, Vignes-Lebbe R, Legendre F. 2017 Taxonomic bias in biodiversity data and societal preferences. Sci. Rep. 7, 9132. (doi:10.1038/s41598-017-09084-6)
8. Meyer C, Kreft H, Guralnick R, Jetz W. 2015 Global priorities for an effective information basis of biodiversity distributions. Nat. Commun. 6, 8221. (doi:10.1038/ncomms9221)
9. Baranzelli MC, Villalobos F, Cordier JM, Nori J. 2023 Knowledge shortfalls’ interactions shadow our perception of species’ exposure to human threats. Biol. Conserv. 282, 110069.
(doi:10.1016/j.biocon.2023.110069)
10. Revell LJ. 2010 Phylogenetic signal and linear regression on species data. Methods Ecol. Evol. 1, 319–329. (doi:10.1111/j.2041-210X.2010.00044.x)
11. Diniz-Filho JAF, Loyola RD, Raia P, Mooers AO, Bini LM. 2013 Darwinian shortfalls in biodiversity conservation. Trends Ecol. Evol. (Amst.) 28, 689–695. (doi:10.1016/j.tree.2013.09.
003)
12. Hinchliff CE et al. 2015 Synthesis of phylogeny and taxonomy into a comprehensive tree of life. Proc. Natl Acad. Sci. USA 112, 12764–12769. (doi:10.1073/pnas.1423041112)
13. Rudbeck AV, Sun M, Tietje M, Gallagher RV, Govaerts R, Smith SA, Svenning J, Eiserhardt WL. 2022 The Darwinian shortfall in plants: phylogenetic knowledge is driven by range size.
Ecography 2022, e06142. (doi:10.1111/ecog.06142)
14. Soares BE, Nakamura G, Freitas TMS, Richter A, Cadotte M. 2023 Quantifying and overcoming Darwinian shortfalls to conserve the fish tree of life. Biol. Conserv. 285, 110223. (doi:
10.1016/j.biocon.2023.110223)
15. Šmíd J. 2022 Geographic and taxonomic biases in the vertebrate tree of life. J. Biogeogr. 49, 2120–2129. (doi:10.1111/jbi.14491)
16. Colston TJ, Kulkarni P, Jetz W, Pyron RA. 2020 Phylogenetic and spatial distribution of evolutionary diversification, isolation, and threat in turtles and crocodilians (non-avian
archosauromorphs). BMC Evol. Biol. 20, 81. (doi:10.1186/s12862-020-01642-3)
17. Jetz W, Thomas GH, Joy JB, Hartmann K, Mooers AO. 2012 The global diversity of birds in space and time. Nature New Biol. 491, 444–448. (doi:10.1038/nature11631)
18. Jetz W, Pyron RA. 2018 The interplay of past diversification and evolutionary isolation with present imperilment across the amphibian tree of life. Nat. Ecol. Evol. 2, 850–858. (doi:
10.1038/s41559-018-0515-5)
19. Tonini JFR, Beard KH, Ferreira RB, Jetz W, Pyron RA. 2016 Fully-sampled phylogenies of squamates reveal evolutionary patterns in threat status. Biol. Conserv. 204, 23–31. (doi:10.
1016/j.biocon.2016.03.039)
20. Upham NS, Esselstyn JA, Jetz W. 2019 Inferring the mammal tree: species-level sets of phylogenies for questions in ecology, evolution, and conservation. PLoS Biol. 17, e3000494.
(doi:10.1371/journal.pbio.3000494)
21. Titley MA, Snaddon JL, Turner EC. 2017 Scientific research on animal biodiversity is systematically biased towards vertebrates and temperate regions. PLoS One 12, e0189577. (doi:
10.1371/journal.pone.0189577)
22. Scornavacca C, Delsuc F, Galtier N. 2020 Phylogenetics in the genomic era. See https://hal.science/hal-02535070v3/.
23. Gaston KJ. 1991 Body size and probability of description: the beetle fauna of Britain. Ecol. Entomol. 16, 505–508. (doi:10.1111/j.1365-2311.1991.tb00242.x)
24. Meiri S. 2018 Traits of lizards of the world: variation around a successful evolutionary design. Glob. Ecol. Biogeogr. 27, 1168–1172. (doi:10.1111/geb.12773)
25. Moura MR, Jetz W. 2021 Shortfalls and opportunities in terrestrial vertebrate species discovery. Nat. Ecol. Evol. 5, 631–639. (doi:10.1038/s41559-021-01411-5)
26. Roll U, Mittermeier JC, Diaz GI, Novosolov M, Feldman A, Itescu Y, Meiri S, Grenyer R. 2016 Using Wikipedia page views to explore the cultural importance of global reptiles. Biol.
Conserv. 204, 42–50. (doi:10.1016/j.biocon.2016.03.037)
27. Guedes JJM, Moura MR, Alexandre F. Diniz‐Filho J. 2023 Species out of sight: elucidating the determinants of research effort in global reptiles. Ecography 2023, 1–14. (doi:10.1111/
ecog.06491)
28. Mammola S et al. 2023 Drivers of species knowledge across the tree of life. Elife 12, RP88251. (doi:10.7554/eLife.88251)
29. Chutipong W, Lynam AJ, Steinmetz R, Savini T, Gale GA. 2014 Sampling mammalian carnivores in western thailand: issues of rarity and detectability. Raffles Bull. Zool. 62, 521–535.
https://lkcnhm.nus.edu.sg/publications/raffles-bulletin-of-zoology/volumes/volume-62/
30. Ribeiro-Júnior MA, Gardner TA, Ávila-Pires TCS. 2008 Evaluating the effectiveness of herpetofaunal sampling techniques across a gradient of habitat change in a tropical forest
landscape. J. Herpetol. 42, 733. (doi:10.1670/07-097R3.1)
31. Colli GR, Fenker J, Tedeschi LG, Barreto-Lima AF, Mott T, Ribeiro SLB. 2016 In the depths of obscurity: knowledge gaps and extinction risk of Brazilian worm lizards (Squamata,
Amphisbaenidae). Biol. Conserv. 204, 51–62. (doi:10.1016/j.biocon.2016.07.033)
32. Johnson KR, Owens IFP, Global Collection Group. 2023 A global approach for natural history museum collections. Science 379, 1192–1194. (doi:10.1126/science.adf6434)
33. GBIF.org. 2022 Global Biodiversity Information Facility. See https://www.gbif.org (accessed 17 April 2024).
34. Callaghan CT, Borda-de-Água L, van Klink R, Rozzi R, Pereira HM. 2023 Unveiling global species abundance distributions. Nat. Ecol. Evol. 7, 1600–1609. (doi:10.1038/s41559-023-
02173-y)
35. Fontaine B, Perrard A, Bouchet P. 2012 21 years of shelf life between discovery and description of new species. Curr. Biol. 22, R943–4. (doi:10.1016/j.cub.2012.10.029)
36. Guedes JJM, Feio RN, Meiri S, Moura MR. 2020 Identifying factors that boost species discoveries of global reptiles. Zool. J. Linn. Soc. 190, 1274–1284. (doi:10.1093/zoolinnean/
zlaa029)
37. Meyer C, Jetz W, Guralnick RP, Fritz SA, Kreft H. 2016 Range geometry and socio-economics dominate species-level biases in occurrence information. Glob. Ecol. Biogeogr. 25, 1181–
1193. (doi:10.1111/geb.12483)
38. Diniz‐Filho JAF, Bastos RP, Rangel TFLVB, Bini LM, Carvalho P, Silva RJ. 2005 Macroecological correlates and spatial patterns of anuran description dates in the Brazilian cerrado.
Glob. Ecol. Biogeogr. 14, 469–477. (doi:10.1111/j.1466-822X.2005.00165.x)
39. Borregaard MK, Rahbek C. 2010 Causality of the relationship between geographic distribution and species abundance. Q. Rev. Biol. 85, 3–25. (doi:10.1086/650265)
40. Jarić I, Bellard C, Courchamp F, Kalinkat G, Meinard Y, Roberts DL, Correia RA. 2020 Societal attention toward extinction threats: a comparison between climate change and
biological invasions. Sci. Rep. 10, 11085. (doi:10.1038/s41598-020-67931-5)
41. Rodrigues ASL, Gray CL, Crowter BJ, Ewers RM, Stuart SN, Whitten T, Manica A. 2010 A global assessment of amphibian taxonomic effort and expertise. Bioscience 60, 798–806. 8
(doi:10.1525/bio.2010.60.10.6)
42. Moura MR, Costa HC, Peixoto MA, Carvalho ALG, Santana DJ, Vasconcelos HL. 2018 Geographical and socioeconomic determinants of species discovery trends in a biodiversity

royalsocietypublishing.org/journal/rsbl Biol. Lett. 20: 20240216

hotspot. Biol. Conserv. 220, 237–244. (doi:10.1016/j.biocon.2018.01.024)
43. Vale MM, Jenkins CN. 2012 Across‐taxa incongruence in patterns of collecting bias. J. Biogeogr. 39, 1744–1748. (doi:10.1111/j.1365-2699.2012.02750.x)
44. Ducatez S, Lefebvre L. 2014 Patterns of research effort in birds. PLoS One 9, e89955. (doi:10.1371/journal.pone.0089955)
45. Goodwin ZA, Muñoz-Rodríguez P, Harris DJ, Wells T, Wood JRI, Filer D, Scotland RW. 2020 How long does it take to discover a species? Syst. Biodivers. 18, 784–793. (doi:10.1080/
14772000.2020.1751339)
46. Raxworthy CJ, Smith BT. 2021 Mining museums for historical DNA: advances and challenges in museomics. Trends Ecol. Evol. (Amst.) 36, 1049–1060. (doi:10.1016/j.tree.2021.07.
009)
47. Fahey AL, Ricklefs RE, Dewoody JA. 2014 DNA-based approaches for evaluating historical demography in terrestrial vertebrates. Biol. J. Linn. Soc. Lond 112, 367–386. (doi:10.1111/
bij.12259)
48. Moura MR et al. 2024 A phylogeny-informed characterisation of global tetrapod traits addresses data gaps and biases. PLoS Biol. 22, e3002658. (doi:10.1101/2023.03.04.531098)
49. Oliveira BF, Scheffers BR. 2019 Vertical stratification influences global patterns of biodiversity. Ecography 42, 249–249. (doi:10.1111/ecog.03636)
50. González-Del-Pliego P, Freckleton RP, Edwards DP, Koo MS, Scheffers BR, Pyron RA, Jetz W. 2019 Phylogenetic and trait-based prediction of extinction risk for data-deficient
amphibians. Curr. Biol. 29, 1557–1563.(doi:10.1016/j.cub.2019.04.005)
51. IUCN. The IUCN red list of threatened species. See https://www.iucnredlist.org (accessed 19 March 2024).
52. Jetz W, McPherson JM, Guralnick RP. 2012 Integrating biodiversity distribution knowledge: toward a global map of life. Trends Ecol. Evol. (Amst.) 27, 151–159. (doi:10.1016/j.tree.
2011.09.007)
53. Marsh CJ et al. 2022 Expert range maps of global mammal distributions harmonised to three taxonomic authorities. J. Biogeogr. 49, 979–992. (doi:10.1111/jbi.14330)
54. Roll U et al. 2017 The global distribution of tetrapods reveals a need for targeted reptile conservation. Nat. Ecol. Evol. 1, 1677–1682. (doi:10.1038/s41559-017-0332-2)
55. Hurlbert AH, Jetz W. 2007 Species richness, hotspots, and the scale dependence of range maps in ecology and conservation. Proc. Natl Acad. Sci. USA 104, 13384–13389. (doi:10.
1073/pnas.0704469104)
56. Amatulli G, Domisch S, Tuanmu MN, Parmentier B, Ranipeta A, Malczyk J, Jetz W. 2018 A suite of global, cross-scale topographic variables for environmental and biodiversity
modeling. Sci. Data 5, 180040. (doi:10.1038/sdata.2018.40)
57. Klein Goldewijk K, Beusen A, Doelman J, Stehfest E. 2017 Anthropogenic land use estimates for the Holocene - HYDE 3.2. Earth Syst. Sci. Data 9, 927–953. (doi:10.5194/essd-9-927-
2017)
58. Kier G, Barthlott W. 2001 Measuring and mapping endemism and species richness: a new methodological approach and its application on the flora of Africa. Biodivers. Conserv. 10,
1513–1529. (doi:10.1023/A:1011812528849)
59. Chamberlain S, Barve V, Mcglinn D, Oldoni D, Desmet P, Geffert L et al. rgbif: Interface to the Global Biodiversity Information Facility API. R package version 3.7.8. See https://cran.r-
project.org/package=rgbif (accessed 13 November 2023).
60. Chamberlain S. 2020 Rredlist: "IUCN” red list client. R package version. See https://cran.r-project.org/package=rredlist.
61. Zuur AF, Ieno EN, Walker NJ, Saveliev AA, Smith GM. 2009 Statistics for biology and health. Berlin, Germany: Springer.
62. Bolker BM, Brooks ME, Clark CJ, Geange SW, Poulsen JR, Stevens MHH, White JSS. 2009 Generalized linear mixed models: a practical guide for ecology and evolution. Trends Ecol.
Evol. (Amst.) 24, 127–135. (doi:10.1016/j.tree.2008.10.008)
63. Harrison XA, Donaldson L, Correa-Cano ME, Evans J, Fisher DN, Goodwin CED, Robinson BS, Hodgson DJ, Inger R. 2018 A brief introduction to mixed effects modelling and multi-
model inference in ecology. PeerJ 6, e4794. (doi:10.7717/peerj.4794)
64. Castiglione S, Melchionna M, Profico A, Sansalone G, Modafferi M, Mondanaro A, Wroe S, Piras P, Raia P. 2022 Human face‐off: a new method for mapping evolutionary rates on
three‐dimensional digital models. Palaeontology 65, 1–10. (doi:10.1111/pala.12582)
65. Castiglione S, Serio C, Mondanaro A, Melchionna M, Raia P. 2022 Fast production of large, time calibrated, informal supertrees with tree.merger. Palaeontology 65, 1–11. (doi:10.
1111/pala.12588)
66. Bolker B, Butler M, Cowan P, de VD, Eddelbuettel D, Holder M et al. 2020 phylobase: Base package for phylogenetic structures and comparative data. See https://cran.r-project.org/
package=phylobase.
67. Keck F, Rimet F, Bouchez A, Franc A. 2016 Phylosignal: an R package to measure, test, and explore the phylogenetic signal. Ecol. Evol. 6, 2774–2780. (doi:10.1002/ece3.2051)
68. George D, Mallery P. 2010 SPSS for windows step by step: a simple guide and reference, 17.0 update, 10th ed. Boston, MA: Allyn & Bacon.
69. Kutner MH, Nachtsheim CJ, Neter J, Li W. 2005 Applied linear statistical models. New York, NY: McGraw-Hill Irwin.
70. Dinerstein E et al. 2017 An ecoregion-based approach to protecting half the terrestrial realm. Bioscience 67, 534–545. (doi:10.1093/biosci/bix014)
71. Hartig F. 2022 DHARMa: residual diagnostics for hierarchical (multi-level/mixed) regression models. See https://cran.r-project.org/package=DHARMa.
72. Lüdecke D, Ben-Shachar M, Patil I, Waggoner P, Makowski D. Performance: an R package for assessment, comparison and testing of statistical models. JOSS 6, 3139. (doi:10.21105/
joss.03139)
73. Bates D, Mächler M, Bolker B, Walker S. Fitting linear mixed-effects models using Lme4. J. Stat. Softw. 67, 1–48. (doi:10.18637/jss.v067.i01)
74. Naimi B, Hamm NAS, Groen TA, Skidmore AK, Toxopeus AG. 2014 Where is positional uncertainty a problem for species distribution modelling. Ecography 37, 191–203. (doi:10.
1111/j.1600-0587.2013.00205.x)
75. R Core Team. 2023 R: A language and environment for statistical computing. Vienna, Austria. See https://www.r-project.org/.
76. Nakamura G, Richter A, Soares BE. 2021 Fishphylomaker: an R package to generate phylogenies for ray-finned fishes. Ecol. Inform. 66, 101481. (doi:10.1016/j.ecoinf.2021.101481)
77. Faith DP. 1992 Conservation evaluation and phylogenetic diversity. Biol. Conserv. 61, 1–10. (doi:10.1016/0006-3207(92)91201-3)
78. Harvey PH, Pagel MD. 1991 The comparative method for studying adaptation. In The comparative method in evolutionary biology, pp. 1–34. Oxford, UK: Oxford university press. (doi:
10.1093/oso/9780198546412.001.0001). See https://academic.oup.com/book/53162.
79. Mouquet N et al. 2012 Ecophylogenetics: advances and perspectives. Biol. Rev. Camb. Philos. Soc. 87, 769–785. (doi:10.1111/j.1469-185X.2012.00224.x)
80. Nürk NM et al. 2020 Diversification in evolutionary arenas—assessment and synthesis. Ecol. Evol. 10, 6163–6182. (doi:10.1002/ece3.6313)
81. Weedop KB, Mooers AØ, Tucker CM, Pearse WD. 2019 The effect of phylogenetic uncertainty and imputation on. Anim. Conserv. 22, 527–536. (doi:10.1111/acv.12495)
82. Enquist BJ et al. 2019 The commonness of rarity: global and future distribution of rarity across land plants. Sci. Adv. 5, eaaz0414. (doi:10.1126/sciadv.aaz0414)
83. Hughes AC, Orr MC, Ma K, Costello MJ, Waller J, Provoost P, Yang Q, Zhu C, Qiao H. 2021 Sampling biases shape our view of the natural world. Ecography 44, 1259–1269. (doi:10. 9
1111/ecog.05926)
84. Rocha LA et al. 2014 Specimen collection: an essential tool. Science 344, 814–815. (doi:10.1126/science.344.6186.814)

royalsocietypublishing.org/journal/rsbl Biol. Lett. 20: 20240216

85. Uribe SV, Estades CF. 2014 Reptiles in Monterey pine plantations of the coastal range of central Chile. Rev. Chil. de Hist. Nat. 87, 1–8. (doi:10.1186/s40693-014-0025-z)
86. Henderson RW, Powell R, Martín J, Lopez P. 2016 Arboreal and fossorial reptiles. In Reptile ecology and conservation: a handbook of techniques, pp. 139–153. Oxford, UK: Oxford
University Press. (doi:10.1093/acprof:oso/9780198726135.001.0001)
87. Einoder LD, Southwell DM, Lahoz-Monfort JJ, Gillespie GR, Fisher A, Wintle BA. 2018 Occupancy and detectability modelling of vertebrates in northern Australia using multiple
sampling methods. PLoS One 13, e0203304. (doi:10.1371/journal.pone.0203304)
88. Bebber DP et al. 2010 Herbaria are a major frontier for species discovery. Proc. Natl Acad. Sci. USA 107, 22169–22171. (doi:10.1073/pnas.1011841108)
89. Bickford D, Lohman DJ, Sodhi NS, Ng PKL, Meier R, Winker K, Ingram KK, Das I. 2007 Cryptic species as a window on diversity and conservation. Trends Ecol. Evol. (Amst.) 22, 148–
155. (doi:10.1016/j.tree.2006.11.004)
90. Parsons DJ, Pelletier TA, Wieringa JG, Duckett DJ, Carstens BC. 2022 Analysis of biodiversity data suggests that mammal species are hidden in predictable places. Proc. Natl Acad. Sci.
USA 119, e2103400119. (doi:10.1073/pnas.2103400119)
91. Streicher JW, Wiens JJ, Jocqué M, García-Vázquez UO, Smith EN. 2023 Molecular phylogenetic analyses reveal both underestimation and overestimation of species diversity in
northern rain frogs (Craugastor). J. Vertebr. Biol. 72. (doi:10.25225/jvb.23072)
92. Brown JH, LomolinoMV. 1998 Biogeography, p. 691, 2nd edn. Sunderland, MA: Sinauer Press.
93. Isaac NJB, Turvey ST, Collen B, Waterman C, Baillie JEM. 2007 Mammals on the EDGE: conservation priorities based on threat and phylogeny. PLoS One 2, e296. (doi:10.1371/journal.
pone.0000296)
94. Gumbs R et al. 2023 The Edge2 protocol: advancing the prioritisation of evolutionarily distinct and globally endangered species for practical conservation action. PLoS Biol. 21,
e3001991. (doi:10.1371/journal.pbio.3001991)
95. Guedes JJM, Diniz-Filho JAF, Moura MR. 2024 Data from: Macroecological correlates of darwinian shortfalls across terrestrial vertebrates. Dryad Digital Repository. (doi:10.5061/
dryad.wdbrv15ww)
96. Guedes JJM, Diniz-Filho JAF, Moura M. 2024 Data from: Macroecological correlates of Darwinian shortfalls across terrestrial vertebrates. Figshare. (doi:10.6084/m9.figshare.c.
7358195.v1)