Syst. Biol.

63(6):1010–1017, 2014
© The Author(s) 2014. Published by Oxford University Press, on behalf of the Society of Systematic Biologists. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
DOI:10.1093/sysbio/syu061
Advance Access publication August 19, 2014

Reconciling Hierarchical Taxonomy with Molecular Phylogenies

BEN G. HOLT1,∗ AND KNUD ANDREAS JØNSSON1,2
1 Department of Life Sciences, Imperial College London, Silwood Park Campus, Ascot SL5 7PY; and 2 Department of Life Sciences, Natural History Museum,
Cromwell Road, London SW7 5BD, UK
∗ Correspondence to be sent to: Department of Life Sciences, Imperial College London, Silwood Park campus, Ascot SL5 7PY, UK;
E-mail: b.holt@imperial.ac.uk.

Received 26 February 2014; reviews returned 12 August 2014; accepted 13 August 2014
Associate Editor: Frank Anderson

Downloaded from http://sysbio.oxfordjournals.org/ at New York University on October 17, 2014

Taxonomy has a history dating back to Aristotle evolution, and conservation often attempt to cover a
(350BC) and has facilitated a wide range of developments particular hierarchical rank (for some examples, see
in the biological sciences. Linnaeus’ Systema Naturae Jønsson et al. 2010; Derryberry et al. 2011; Condamine
assumed that organisms were static creations of et al. 2012; Holt et al. 2013). These approaches all
God and formulated the hierarchical framework of implicitly assume that hierarchical ranks are to some
classification that we currently use. Today we know extent comparable across taxa and some studies even
that organisms continuously evolve and it is generally make the assumption explicit by using these categories
accepted that these hierarchies are arbitrary constructs as units of study (e.g., Lagomarcino and Miller 2012;
(Coyne and Orr 2004). The arbitrary nature of these Rueda et al. 2013). Although the extensive use of
higher taxonomic ranks does not prevent their practical current taxonomic rankings demonstrates both their
use with regards to cataloguing, and communicating practical benefits and scientific importance, the fact
about, biological diversity, provided such arbitrary that no attempt has been made to delimit higher
decisions are made on a consistent basis. However, rankings consistently is cause for concern (Hennig 1966;
current taxonomic hierarchies are suggested to be de Queiroz and Gauthier 1992; Avise and Liu 2011).
inconsistent, both between and within major clades We intuitively assume that hierarchical ranks reflect
(Avise and Liu 2011), reflecting the work of systematists similar temporal evolutionary histories across different
with diverging views regarding the constitution of taxa, but this is commonly not the case. For example,
higher taxonomic ranks. In this study, we present the age of orders may vary by more than 400 myr
a comprehensive analysis of the overall levels of between vertebrates and invertebrates (Avise and Liu
consistency within current higher taxonomic ranks 2011).
using dated phylogenies for all bird and mammal Heated debates have proved that it is not a trivial
species. task to produce taxonomies that convey information
Building on work by Hennig (1966), and proposed about the evolutionary history of organisms on Earth
by Avise and Johns (1999), “temporal banding” (Hennig 1966; Mayr 1974; de Queiroz and Cantino 2001).
provides an opportunity to assess the consistency These debates have centered on the evolutionary theory
of ancestral relationships within and among higher that should underpin the assignment of taxonomic
taxonomic ranks, as well as a practical solution ranks, with the two main competing views focusing
to these temporal inconsistencies. Temporal banding on either the identification of lineages entering new
standardizes taxonomic ranks by “cutting” dated adaptive zones or the identification of clades (Mayr
phylogenies at specified points in time and applying 1974). Although both of these views represent a valid
this concept requires comprehensive dated phylogenies, approach to rank taxonomic delimitation in theory, the
which has so far limited the application of the approach. practical application of these concepts requires relevant
Here, we apply temporal banding to two vertebrate data for the focal organisms, as well as an analytical
classes (birds and mammals) and produce standardized approach that can use such data to delimit taxonomies
error metrics to compare the consistency of existing in a systematic manner. In the absence of such an
taxonomic ranks within and between the two classes. We objective analytical approach, higher taxonomic ranks
discuss the implications of error within the phylogenies cannot be expected to represent comparable biological
used, as well as the practicalities of the temporal banding units. Taxonomic inconsistencies have the potential to
approach. bias the results of scientific studies, as well as adversely
Taxonomic inconsistencies have implications for our influencing scientific prioritization and comparisons
view on the uniqueness of organisms. Studies of ecology, across studies. Consequently, new opportunities to

1010

[17:58 9/10/2014 Sysbio-syu061.tex] Page: 1010 1010–1017

 2014 POINTS OF VIEW 1011

quantify the full extent of such taxonomic inconsistencies Taxonomic Clade Ages
need to be explored. The level of divergence (referred to as “clade
The recent emergence of complete dated species-level age”) within taxonomic groups was calculated as
DNA phylogenies for large well-studied animal groups the age, in millions of years ago (ma), of the
such as birds (Jetz et al. 2012) and mammals (Bininda- oldest phylogenetic node connecting the immediately
Emonds et al. 2007; Fritz et al. 2009) has greatly benefited subordinate taxonomic groups within a clade (i.e.,
the study of the natural world. However, these new species within genera, genera within families, and
hypotheses have not yet been used to systematically families within orders). Monotypic taxonomic groups
review the classifications that ornithologists and (with only a single immediately subordinate taxonomic
mammalogists use on a daily basis. Analyzing levels group) were not assigned a clade age and are therefore
of temporal divergence within these large phylogenies not included within this specific aspect of the study.
allows temporal inconsistencies (i.e., inconsistent levels Note that this approach differs subtly, from the approach
of phylogenetic divergence among taxonomic groups) known as “crown age,” which is calculated as the
to be fully quantified within major taxonomic ranks. age of the node representing the most recent common
Temporal divergence is often an important consideration

Downloaded from http://sysbio.oxfordjournals.org/ at New York University on October 17, 2014

ancestor of all species within a clade. We rejected this
with modern taxonomic classifications and subjective approach because it artificially exaggerates taxonomic
interpretations of morphological and anatomical data inconsistencies by unfairly assigning young ages to
to determine taxonomic ranks have clearly led to monotypic families and orders. For example, for a family
major temporal inconsistencies within ranks. This that contains only one genus, but multiple species, both
traditional focus on morphology has been suggested to the family and the genus would have the same crown age
inhibit taxonomic objectivity due to the lack of clear despite the fact that these species would only be expected
homologous features among major groups (Avise and to show genus-level divergence.
Johns 1999).
The goal of this study is to evaluate the consistency
of current taxonomic rankings. We take advantage of New Phylogeny-Based Taxonomies
these large complete, dated phylogenies and examine
the extent of temporal inconsistency across bird and For analytical purposes, new hierarchical taxonomic
mammal genera, families, and orders. Specifically, we rankings (orders, families, and genera) were produced
(i) produce the first full temporal banding classifications by cutting the phylogenies at the age that returned
for these groups, (ii) explore the extent of current the same number of groups as found in the original
temporal inconsistencies in the ages of hierarchical rankings. These arbitrary cut-off ages allow for direct
clades, and finally (iii) test for phylogenetic bias in these comparisons with the original classifications and we are
inconsistencies, that is, are specific types of temporal neither supporting nor rejecting the wider use of these
inconsistency more commonly found toward the base cut-off ages to produce taxonomic hierarchies for these
or the tips of the trees? clades. Since the phylogenies are not fully resolved (the
bird phylogeny contains some polytomies, in the form
of zero length phylogenetic branches), it was not always
possible to return the exact number of groups within
METHODS ranks as found in the original classification. In these
cases, we used the cut-off age that returned the closest
Phylogenies Used possible number of groups within ranks to the original
For birds, we used the taxonomy and phylogeny from number in the taxonomy.
Jetz et al. (2012). These data consisted of 9993 species,
divided into 40 orders, 194 families, and 2091 genera. The
bird phylogeny is based on molecular data from 6670 Consistency among Taxonomic Schemes
species, with the remaining species added according We calculated error scores for each hierarchical
to taxonomy. In order to investigate the potential for taxonomic rank (Fig. 1). This method measures whether
bias created by the taxonomic addition of these species taxonomic groups have been either (i) split from their
without data, we repeated our analysis using only sister clade when they should not have been (i.e., the
the 6670 species for which molecular data existed. In split is more recent than the standardized cut-off age) or
addition, because the bird taxonomy used by Jetz et al. (ii) lumped with their sister clade when they should not
(2012) was not based on a single existing taxonomy but have been (i.e., the split with the sister clade is older than
an amalgamation of several, we also ran our analysis the standardized cut-off age). In both cases “temporal
using the International Ornithological Committee (IOC) error” is measured in millions of years (myr) from the
taxonomy (Gill and Donsker 2013). For mammals, we erroneous taxonomic split and the standardized cut-off
used the taxonomy of Wilson and Reeder (2005) and the age, with taxonomic clades that are split too recently
phylogeny of Bininda-Emonds et al. (2007) with updates giving negative values and taxonomic clades that contain
according to Fritz et al. (2009). These mammal data splits older than the cut-off age giving positive values.
consisted of 5020 species, divided into 29 orders, 148 The mean amount of absolute error was calculated across
families, and 1188 genera. all groups.

[17:58 9/10/2014 Sysbio-syu061.tex] Page: 1011 1010–1017

 1012 SYSTEMATIC BIOLOGY VOL. 63

Downloaded from http://sysbio.oxfordjournals.org/ at New York University on October 17, 2014

FIGURE 1. Graphical representation of temporal “error scores.” In all cases, the phylogeny is cut to produce four clades (the same as the number
of genera according to the existing taxonomy). a) The empirical error is calculated (Genera: Platysmurus, Temnurus Crypsirina, and Dendrocitta).
According to the “cut-off age” (dotted green line), the genera Platysmurus and Temnurus should be lumped (indicated by the red color). Dendrocitta
formosae should be split from its congeners (indicated in blue). Crypsirina is consistent (indicated in gray) and consequently produces no error.
The total error score is calculated as the sum of the absolute distances (in myr) between the node that causes temporal inconsistency and the
cut-off age All possible random splits are shown as (b to e), including the one that matches the empirical case (d). For the main analysis, random
splits were generated 1000 times and the empirical result was compared with the mean random result to produce standardized error scores.

These raw scores are not directly comparable because is randomly distributed across phylogenetic trees or if it
expected levels of error differ among study groups and is clustered toward the base or the tips. We conducted
among taxonomic ranks. “Standardized error scores” Spearman correlation tests to consider potential
were therefore produced for each taxonomic rank within correlations between taxonomic clade phylogenetic
each taxonomic group. These scores compare the total position and clade temporal error score, for all three
amount of error to random expectations, resulting in hierarchical ranks considered and for both birds and
a metric for which zero represents a perfect score (a mammals. Taxonomic clade phylogenetic position was
completely consistent taxonomy), and one represents quantified as root node distance, that is, the number of
a score that is equal to the mean random expectation. nodes between the root and ancestral node of the clade,
Random expectations were generated by randomly with clades having low root node distances referred
dividing phylogenies into the required number of to as being more basal and those with higher scores
monophyletic groups (Fig. 1). For example, for random referred to as being more distal. Clade temporal error
bird orders, all bird species were randomly divided into scores were calculated as explained in the “Consistency
40 monophyletic groups in accordance with the number among taxonomic schemes” section.
of orders in the bird taxonomy. Species were divided into
random clades by randomly splitting the phylogenetic
nodes, starting with the root, until the required number RESULTS
of clades was produced. Nodes were randomly selected
for splitting in a weighted manner that ensured that For birds, the new phylogenetically delineated orders,
all possible random results were equally possible. This families, and genera were generated by cutting the
process was repeated 1000 times and the mean random phylogeny at 65, 37.5, and 11.405 ma, respectively
score calculated. Standardized error scores were then (Supplementary Material S1, (available from http://
calculated by dividing the total observed error by the dx.doi.org/10.5061/dryad.qd3pd). Standardized error
mean total random error. All analyses were performed scores for the three bird taxonomic ranks show
using the R statistical analysis software (R Core Team that the taxonomy is most consistent at the order
2013), using the packages ape, foreach and picante level and least consistent at the genus level (Fig. 2).
(Paradis et al. 2004; Kembel et al. 2010; Analytics and Overall 695 of the 2325 taxonomic clades were
Weston 2013). completely consistent between the original and revised
classifications, with roughly half of consistent clades
being monotypic. Traditional hierarchical taxonomic
clade age distributions overlapped considerably (Fig. 3a)
Phylogenetic Bias in Taxonomic Inconsistencies and clades varied substantially in their amount of
We investigated whether particular parts of temporal error (Fig. 4a–c; Supplementary Material
phylogenies are more prone to inconsistent taxonomic S2). The most extreme comparison was between the
treatment. Specifically, we ask whether temporal error order Casuariiformes, “emus and cassowaries” (clade

[17:58 9/10/2014 Sysbio-syu061.tex] Page: 1012 1010–1017

 2014 POINTS OF VIEW 1013

100
80 1.0 Mean Random

0.8

Standardised Error (points)

% consistent (bars)

60

0.6
40

0.4

Downloaded from http://sysbio.oxfordjournals.org/ at New York University on October 17, 2014

20

0.2

0 Perfect
0

Genera Families Orders Genera Families Orders

Birds Mammals

FIGURE 2. Consistency of current higher taxonomic classifications for birds and mammals, for genera, families, and orders. As measured by
percentage of groups that are identical to a group produced by a temporal banding approach (bars) or by standardized error scores (points), for
which zero = no error and one = mean random error. Standardized error scores are the most appropriate metric for comparing study groups
or hierarchical ranks. See the “Methods” section for analytical protocol for producing standardized error scores.

age = 22.31 ma), and the genera Caprimulgus and the order Dasyuromorphia: carnivorous marsupials
Eurostopodus, both nightjars, with clade ages of 55.13 ma. (clade age = 30.5 ma), and the genus Abrocoma (clade
Our two repeated versions of the bird analysis, one using age = 45.3 ma), which is within the chinchilla rat family,
the data-only phylogeny and the other using the IOC Abrocomidae.
taxonomic classification, both produced highly similar For both birds and mammals, families and genera
results to the main analysis (Supplementary Material S3). showed a significant negative correlation between
For mammals, the new phylogenetically based orders, root node distances and clade error scores (birds
families, and genera were generated by cutting the rs = −0.54 and −0.46; mammals: rs = −0.49 and −0.43,
phylogeny at 75.7, 44.2, and 17.05 ma, respectively both P<0.001; Fig. 3b,c,e,f). This result suggests that
(Supplementary Material S1). Due to polytomies, it more basal taxonomic clades are more likely to be split
was not possible to find cut-off ages returning the under temporal banding and more distal clades are more
same number of clades for families and genera as likely to be lumped. Orders did not show any significant
suggested by Wilson and Reeder (2005). The selected correlation for either birds or mammals (birds:
cut-off ages returned the closest possible number rs = −0.22, P = 0.18; mammals: rs = −0.07, P = 0.7;
of clades, with 10 more genera (1198) and 2 more Fig. 3a,d). Analytical code is available as Supplementary
families (150). Mammalian standardized error scores Material S4.
again show that the taxonomy is most consistent
at the order level and least consistent at the genus
level (Fig. 2). Within hierarchical taxonomic ranks, the DISCUSSION
mammal taxonomy was more consistent than the bird Quantitative analyses presented here demonstrate
taxonomy (Fig. 2). Overall, 533 of the 1365 original that the current bird taxonomy is more inconsistent
mammal clades were completely consistent with a than that of mammals within the three taxonomic
clade in the temporal banding scheme, with more ranks considered. Within hierarchical taxonomic ranks,
than half of consistent clades being monotypic. The mammal groups are consistently older than bird
traditional mammal taxonomic clade age distributions groups. For both birds and mammals, the amount
again overlapped strikingly across taxonomic ranks of temporal inconsistency among clades decreases as
(Fig. 3b) and clades again varied substantially in their taxonomic hierarchical ranks increase. Bird genera are
amount of temporal error (Fig. 4d–f; Supplementary particularly inconsistent and only slightly better than
Material S2). The most extreme comparison was between a randomly delineated taxonomy. Analysing only the

[17:58 9/10/2014 Sysbio-syu061.tex] Page: 1013 1010–1017

 1014 SYSTEMATIC BIOLOGY VOL. 63

Downloaded from http://sysbio.oxfordjournals.org/ at New York University on October 17, 2014

FIGURE 3. Frequency density plots for current hierarchical taxonomic clade ages of a) birds and b) mammals following traditional taxonomy
and the new phylogeny-based temporal banding approach, in which time calibrated phylogenies are cut at specific points in time to delineate
consistent hierarchical taxonomic ranks.

FIGURE 4. Phylogenetic bias of taxonomic consistencies for bird and mammal orders, families, and genera. Dashed lines represent the cut-off
ages required to split the phylogenies into the existing number of groups for each taxonomic rank. If a clade is split too recently (i.e., “Should
be lumped”) the phylogenetic lineage is shaded red. If a taxonomic clade is split too long ago (i.e., “Should be split”) the phylogenetic lineage
is shaded blue. In both cases the intensity of color reflects the temporal distance from the dashed cut-off age. Lineages that are consistent with
the temporal banding approach are shaded gray. For birds and mammals, a significant negative correlation exists between the root node age of
clades and clade temporal error for families and genera (birds rs = −0.54 and −0.46; mammals: rs = −0.49 and −0.43, both P < 0.001). For orders
no significant correlation exists between these variables (birds: rs = −0.22, P = 0.18; mammals: rs = −0.07, P = 0.7).

[17:58 9/10/2014 Sysbio-syu061.tex] Page: 1014 1010–1017

 2014 POINTS OF VIEW 1015

part of the bird phylogeny for which molecular data families were lumped. For birds, 12 of the most distal
exists (6670 species) provides congruent results; this third of families (65 families) are split under temporal
suggests that the temporal banding approach may even banding but many more of these families are lumped
be applied to incomplete phylogenies and ultimately (47). The most extreme case of lumping for birds (in
to the whole tree of life. Across current families and terms of number of old families lumped) is represented
genera, basal taxonomic groups are more likely to be split by the existing superfamily Passeroidea, which consists
under a temporal banding approach, with significant of 15 families, 14 of which are lumped into one single
negative correlations between phylogenetic position and family by temporal banding, with the remaining family
temporal error within these taxonomic ranks, for both being the single species family Urocynchramidae. In
study groups. the corresponding example of extreme lumping under
Our results have important implications for bird and mammalian temporal banding, six current families are
mammal studies based on higher taxonomic ranks and, lumped into a single family that is currently recognized
since our results are consistent across the two groups, as the suborder Feliformia.
there is also the potential that similar studies based on Taken together it appears that the coarse delimitations
other groups could be influenced in a similar manner. of birds and mammals into orders have been carried

Downloaded from http://sysbio.oxfordjournals.org/ at New York University on October 17, 2014

Results for orders compare well with null expectations out in a relatively consistent manner by taxonomists,
and are encouraging for current taxonomies. This whereas the more fine-tuned delimitations into families
hierarchical rank shows relatively little error and no and genera have led to a wide variation of temporal
evidence of any phylogenetic bias for splitting or groupings. The total number of groups may play an
lumping. This result suggests that currently observed important role with regards to maintaining consistency
imbalances in the number of species within orders across taxonomic delimitations. For example, it is
may well be a fairly accurate reflection of similar conceivable that a taxonomist can consider all 29 existing
imbalance in ancestral relationships. For example, the groups when evaluating current mammalian orders but
hyper-diverse bird order Passeriformes remains almost maintaining an overview of over a thousand genera
completely intact under temporal banding, with only would be more problematic without formal analytical
2 of the 5966 species being removed (New Zealand techniques.
wrens, Acanthisittidae). For mammals, two hyper- Although we rely on two complete phylogenies for
diverse orders currently exist, Rodentia and Chiroptera. birds and mammals, it is important to be aware that
Although Rodentia is split into five new orders under the underlying phylogenetic data are not perfect. An
temporal banding, one of these new orders still accounts uncertain proportion of the temporal inconsistency
for over a quarter of mammal species. Chiroptera shown may be due to phylogenetic error. Although these
remains fully intact and accounts for over a fifth of large phylogenies represent important achievements,
mammal species. they also represent the first attempts at establishing
Families show a level of consistency that is complete phylogenetic hypotheses at the scale of
intermediate between that shown by orders and genera; classes (Ricklefs and Pagel 2012). Until new phylogenies
however, temporal banding revision at this hierarchical provide some indication of the levels of uncertainty
rank affects more species than at the other two ranks. in these phylogenies, we would not recommend full
The reason for this is that revised families tend to contain taxonomic revisions based on our new taxonomies
more species than families that remain unchanged. As a without further scrutiny. This is particularly relevant
result 82% of bird species and 74% of mammal species at genus level where the influence of species placed
have their families revised under temporal banding. The within the phylogenies without associated data is likely
nature of temporal banding family revisions appears to be greatest. Nevertheless, it should be acknowledged
to be influenced by the position of families within that these phylogenies represent the state-of-the-art
the phylogenies. For birds and mammals, most of the of knowledge of systematic relationships within these
existing families that were split under temporal banding groups and the results of our temporal banding highlight
represent basal lineages, although some exceptions exist many specific lineages that may warrant taxonomic
for birds. The most basal bird family, Tinamidae (the revision. The results for our restricted bird analysis,
sole family within Tinamiformes), is also the most which includes only species placed within the phylogeny
extreme example of temporal banding splitting, leading based on molecular data (6670 species out of 9993
to splitting into two orders (equivalent to the currently species), are similar to the main results, which suggests
recognized subfamilies of Nothurinae and Tinaminae), that the temporal banding approach may be applied to
with four new families each. Notable mammalian incomplete phylogenies.
examples include Didelphidae and Dasypodidae, two of As more data are collected, and phylogenetic methods
the most basal mammalian lineages, with Didelphidae improve, we expect further systematic rearrangements
split into five new families under temporal banding and within birds and mammals. Moreover, because
Dasypodidae split into two families separated by over hierarchical ranks are arbitrarily defined constructs,
70 myr (close to the cut-off age for orders). Very few there is not a single correct number of groups at any
of the most distal mammalian families were split under particular rank. For analytical purposes, we delineated
temporal banding; for example, only 3 of the most distal taxonomic hierarchical ranks using cut-off ages that
third of the families (50 families), whereas 36 of these returned the same number of groups shown by current

[17:58 9/10/2014 Sysbio-syu061.tex] Page: 1015 1010–1017

 1016 SYSTEMATIC BIOLOGY VOL. 63

taxonomies. It is entirely feasible that alternative cut- decisions can be made on a systematic, consistent basis.
off ages may be preferred, perhaps relying on other The temporal banding approach itself is not completely
characters such as morphology and anatomy, to fine free from practical challenges. Without the possibility of
tune the temporal cut-off ages. Unfortunately, it is only molecular data, fossils must still be assigned to lineages
too easy to envision the levels of disagreement among through their morphological characteristics. Following
expert taxonomists on such fine-tuned cut-off ages and, this process however, fossils could be assigned to higher
given the arbitrary nature of these ranks, the approach groups defined by temporal banding ranks in much the
used for our analysis is more practical, consistent, and same manner as they are currently.
objective. Alternatively, existing or future objective Our study demonstrates that the temporal banding
analytical methods may offer an approach that returns approach can produce consistently defined taxonomic
nonarbitrary cut-off ages (Humphreys and Barraclough groups. These groups therefore represent more
2014). meaningful units of comparison when discussing other
The Linnean hierarchical taxonomic system has aspects of interest, such as phenotypic divergence.
proven resilient and continues to have a strong influence Temporal banding hierarchies also provide more
on modern taxonomy. Despite the increasing availability consistent units of study for formal analysis, however,

Downloaded from http://sysbio.oxfordjournals.org/ at New York University on October 17, 2014

of phylogenetic data, and some calls for the abolition in many cases it may be preferable to perform such
of higher taxonomic ranks (de Queiroz and Gauthier analysis using the phylogenies directly rather than
1992), the use of this classification system remains as through a taxonomic classification. Our consistent
prevalent as ever both within science and across the taxonomies should stimulate a re-evaluation of specific
wider public. However, current taxonomic hierarchical current taxonomic groups for birds and mammals,
ranks have not yet been demonstrated to contain and this study shows that a consistent delineation of
comparable groups for any aspect of phenotypic or large organismal ranks is both theoretically achievable
genotypic diversity. Linnaeus’ famous quote: “God and practically feasible. Ultimately, until a consistent
created, Linnaeus organized” captures the essence of delineation of taxonomic ranks is applied to the complete
the vision of Linnaeus’ work and implies that Linnaeus, tree of life, taxonomies will still fail to accurately reflect
humble to creation by God, merely sought to catalog the continuously evolving nature of life on Earth.
the creations of the master. The publication of “On the
Origin of Species” by Charles Darwin (1859) marked the
beginning of a new era in biological thinking. Rather SUPPLEMENTARY MATERIAL
than organisms being static in need of cataloguing
they were considered continuously changing and Data available from the Dryad Digital Repository:
evolving. Such reinterpretation is naturally assumed http://dx.doi.org/10.5061/dryad.qd3pd.
to lead to a taxonomic paradigm shift away from
traditional taxonomic practises and toward an emphasis
on evolutionary processes (e.g., Hennig 1966; de Queiroz FUNDING
1997). Under temporal banding, for a given hierarchical B.G.H. acknowledges the support of Imperial College
rank, a pair of species within a taxonomic group London’s Grand Challenges in Ecosystems and the
must be more closely related than any pair of species Environment Initiative. This work was supported
that are in different taxonomic groups. Our analysis by funding from the People Programme (Marie
demonstrates that this condition is not close to being Curie Actions) of the European Union’s Seventh
met in current bird and mammal taxonomic hierarchies, Framework Programme (FP7/2007–2013) under REA
particularly for families and genera. This temporal grant agreement no. PIEF-GA-2011-300924.
inconsistency seriously impedes the communication of
evolutionary relationships among species, as well as
biological comparative analyses based on taxonomic
hierarchical ranks. ACKNOWLEDGMENTS
The new hierarchical ranks produced in this The authors thank Frank Anderson, Timothy
study convey comparable information about the Barraclough, Pierre-Henri Fabre, Aelys Humphreys,
temporal evolutionary histories of birds and mammals. Trevor Price, Susanne Renner, and James Rosindell
Standardization based on alternative taxonomic for their valuable advice on this study and for their
concepts, such as classifications based on morphology, comments on the article.
ecology, or evolution, might be considered to be a
preferable approach to some but would be practically
highly problematic, if not impossible. For example, REFERENCES
“Evolutionary systematics” as defined by (Mayr 1974)
requires decisions regarding “whether or not it [a Analytics R., Weston S. 2013. foreach: Foreach looping construct
lineage] has entered a new adaptive zone and to what for R. Available from: URL http://CRAN.R-project.org/package=
foreach/
extent it has experienced a major radiation”. Due to Aristotle. 350BC. 1907. The history of animals. [Translation by
the lack of available informative homologous features D.W. Thompson.] London, UK: John Bell. Available from: URL
across the tree of life, it is difficult to envision how such https://archive.org/details/aristotleshistor00arisrich

[17:58 9/10/2014 Sysbio-syu061.tex] Page: 1016 1010–1017

 2014 POINTS OF VIEW 1017

Avise J.C., Johns G.C. 1999. Proposal for a standardized temporal Holt B.G., Lessard J.-P., Borregaard M.K., Fritz S.A., Araújo M.B.,
scheme of biological classification for extant species. Proc. Natl Dimitrov D., Fabre P.-H., Graham C.H., Graves G.R., Jønsson K.A.,
Acad. Sci. U. S. A. 96:7358–7363. Nogués-Bravo D., Wang Z., Whittaker R.J., Fjeldså J., Rahbek C. 2013.
Avise J.C., Liu J.-X. 2011. On the temporal inconsistencies of Linnean An update of Wallace’s zoogeographic regions of the world. Science
taxonomic ranks. Biol. J. Linn. Soc. 102:707–714. 339:74–78.
Bininda-Emonds O.R.P., Cardillo M., Jones K.E., MacPhee R.D.E., Humphreys A.M., Barraclough T.G. 2014. The evolutionary reality of
Beck R.M.D., Grenyer R., Price S.A., Vos R.A., Gittleman J.L., higher taxa in mammals. Proc. Biol. Sci. 281:20132750.
Purvis A. 2007. The delayed rise of present-day mammals. Nature Jetz W., Thomas G.H., Joy J.B., Hartmann K., Mooers A.O. 2012. The
446:507–512. global diversity of birds in space and time. Nature 491:444–448.
Condamine F.L., Sperling F.A.H., Wahlberg N., Rasplus J.-Y., Kergoat Jønsson K.A., Bowie R.C.K., Moyle R.G., Christidis L., Norman J.A.,
G.J. 2012. What causes latitudinal gradients in species diversity? Benz B.W., Fjeldså J. 2010. Historical biogeography of an Indo-Pacific
Evolutionary processes and ecological constraints on swallowtail passerine bird family (Pachycephalidae): different colonization
biodiversity. Ecol. Lett. 15:267–277. patterns in the Indonesian and Melanesian archipelagos.
Coyne J.A., Orr H.A. 2004. Speciation. Sunderland (MA): Sinauer J. Biogeogr. 37:245–257.
Associates, Incorporated Publishers. Kembel S.W., Cowan P.D., Helmus M.R., Cornwell W.K., Morlon
Darwin C. 1859. On the origin of the species by natural selection. H., Ackerly D.D., Blomberg S.P., Webb C.O. 2010. Picante: R
London, UK: Murray. tools for integrating phylogenies and ecology. Bioinformatics
De Queiroz K. 1997. The Linnaean hierarchy and the evolutionization 26:1463–1464.

Downloaded from http://sysbio.oxfordjournals.org/ at New York University on October 17, 2014

of taxonomy, with emphasis on the problem of nomenclature. Aliso Lagomarcino A.J., Miller A.I. 2012. The relationship between genus
15:125–144. richness and geographic area in Late Cretaceous marine biotas:
De Queiroz K., Cantino P.D. 2001. Phylogenetic nomenclature and the epicontinental sea versus open-ocean-facing settings. PLoS One
PhyloCode. Bull. Zool. Nomencl. 58:17. 7:e40472.
De Queiroz K., Gauthier J. 1992. Phylogenetic taxonomy. Annu. Rev. Mayr E. 1974. Cladistic analysis or cladistic classification. Z. Zool. Syst.
Ecol. Syst. 23:449–480. Evol. 12:94–128.
Derryberry E.P., Claramunt S., Derryberry G., Chesser R.T., Cracraft Paradis E., Claude J., Strimmer K. 2004. APE: analyses of phylogenetics
J., Aleixo A., Pérez-Emán J., Remsen J. V, Brumfield R.T. 2011. and evolution in R language. Bioinformatics 20:289–290.
Lineage diversification and morphological evolution in a large-scale R Core Team. 2013. R: A language and environment for statistical
continental radiation: the neotropical ovenbirds and woodcreepers computing. Available from: URL http://www.R-project.org/
(aves: Furnariidae). Evolution 65:2973–2986. Ricklefs R.E., Pagel M. 2012. Evolutionary biology: birds of a feather.
Fritz S.A., Bininda-Emonds O.R.P., Purvis A. 2009. Geographical Nature 491:336–337.
variation in predictors of mammalian extinction risk: big is bad, Rueda M., Rodríguez M.Á., Hawkins B.A. 2013. Identifying global
but only in the tropics. Ecol. Lett. 12:538–549. zoogeographical regions: lessons from Wallace. J. Biogeogr. 40:2215–
Gill F., Donsker D. 2013. IOC World Bird List (v 3.5). Available from: 2225.
URL http://www.worldbirdnames.org/updates/ Wilson D.E., Reeder D.M. 2005. Mammal species of the world: a
Hennig W. 1966. Phylogenetic systematics. University of Illinois taxonomic and geographic reference. Baltimore (MD): John Hopkins
Press. University Press.

[17:58 9/10/2014 Sysbio-syu061.tex] Page: 1017 1010–1017