M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

The Concept of Species in Microbiology

Introduction
In biology, classification is part of a scientific approach aimed at describing and classifying living
beings, in order to better study and understand them. Nevertheless, it remains a concept that faces
certain limitations/exceptions/difficulties in the continuum formed by these different forms of life,
which are constantly evolving.
As early as antiquity, Aristotle (384-322 BC) proposed that the species groups together
individuals that resemble each other, and perpetuate themselves in a stable way (variations being
considered as accidents).
In the 18th century, the Swedish physician and botanist Carl von Linné (1707-1778) established
a so-called "fixist" classification which makes it possible to classify and organize groups of living beings
according to a logical/systematic approach into several categories, namely kingdom, phylum, class,
order, family, genus and species, the fundamental unit. According to Linnaeus, a species is "a set of
individuals that engender, through reproduction, other individuals similar to themselves". His vision is
described as "fixist" because, according to Linnaeus, species, which were created by God, are
immutable. The concept of species then constitutes the smallest taxonomic unit, the basic brick,
grouping together individuals with common morphological and physiological characteristics. The
nomenclature is based on a genus name (written in italics with the first letter in uppercase) followed
by an epithet (in lower case italics) designating the species. Although the concept of species has
evolved, the Linnaeus classification is still used today for macroscopic organisms.
Ex: Mus musculus (the grey mouse, domestic)
The French naturalist Georges-Louis Leclerc de Buffon (1707-1788) introduced in 1749 the
notion of inter-fertilization at the heart of the definition of the species: "One must regard as the same
species that, by means of copulation, perpetuates itself and retains the similarity of that species, and
as different species those that, by the same means, cannot produce anything together". The concept
of species, and more particularly the fixist view of it, was subsequently confronted with the conceptual
difficulties linked to the theories put forward by Jean-Baptiste Lamarck (1744-1829) and Charles
Darwin (1809-1882), according to which species are born, change, transform and are likely to
disappear over time. Very early on, the definition of a species was problematic, as illustrated by the
famous statement by Charles Darwin: "Nor will I discuss here the different definitions that have been
given to the term species. None of these definitions has completely satisfied all naturalists, and yet
each one of them knows vaguely what he means when he speaks of a species. "(The Origin of Species,
1859 - Edition Maspero, 1980, p.45).
The fruitful alliance of the theory of evolution and genetics led to the emergence in the 1940s
of the so-called "synthetic theory of evolution" to which several scientists contributed (Theodosius
Dobzhansky, Julian Huxley, Ernst Mayr, George Gaylord Simpson, Georges Ledyard Stebbins).
Numerous definitions of the species will then be proposed, including evolutionary and ecological
concepts. In 1942, Ernst Mayr stated: "Species are groups of actually and potentially interbreeding
natural populations, which are isolated from other groups". This definition has an ecological dimension
in that it groups together a population of individuals that are phenotypically and ecologically similar to
each other compared to those of another species, and raises the question of the exchange of breeding
stock between populations. In 1951, George Gaylord Simpson also proposed a definition of the species
that put forward the notion of lineage: "a single lineage of ancestor-descendant populations which
maintains its identity from other such lineages and which has its own evolutionary tendencies and
historical fate". Currently, the definition of the species (for macroscopic organisms) generally takes
into account morphological (but difficulties linked to possible sexual dimorphism), phylogenetic and/or
biological criteria (e.g. interbreeding => Mayr's Biological Species Concept). Finally, some definitions
that aim to bring a "universal" character to the species concept highlight the ecological aspects and
the "forces" that maintain the cohesion of the group: a species is a group of organisms whose

1
M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

divergence is limited by a cohesive force; divergence between different species is irreversible; and
different species are ecologically distinct.
Problem: These definitions of the concept of species encounter limits when parthenogenetic
and/or sexless organisms sensu stricto (e.g. prokaryotic and some eukaryotic microorganisms, viruses)
are taken into account. Moreover, many microorganisms have similar morphologies. The notion of
species in Microbiology was therefore immediately difficult to define. In addition, intra- and
interspecific genetic exchanges between microorganisms are important drivers of evolution, and blur
taxonomic boundaries. Interestingly, the notion of lineage remains at the heart of the reflection, and
the best method for tracing it is therefore at the heart of this issue. Note that in microbiology (and
especially in virology), these approaches cannot (or only exceptionally) rely on the availability of fossil
specimens.

Rem: « Typological » = based on the notion of "type" species; « Biological » = based on the notion of
population inter-fertilization ; « Ecological » = based on the notion of ecological niche; « Evolutionary
» = based on the notion of ancestor-descendant lineage (need for a fossil); « Phylogenetic » = group of
organizations that share a common ancestry – From O. Ogunseitan (2004)

Framework taken into account for this essay: archaea, bacteria, eukaryotic microorganisms, viruses
(less treated but a little discussed)
Plan:
- Species concept based on phenotypic traits
- Species concept based on genome analysis
- Towards a species concept integrating the spatio-temporal and ecological dimensions of the microbial
world

1. Species concept based on phenotypic traits

1.1. Definition and examples
The first attempts at so-called "phenotypic" classification of microorganisms were based on
the phenotypic grouping of isolates on the basis of certain morphological (with or without staining),
cultural, physiological and biochemical characteristics, in particular.
E.g.: colony aspects, motility, rod or shell shapes, possible groupings (e.g. tetrad), Gram staining, use
of certain carbohydrates, amino acid metabolism, urea metabolism, anaerobic metabolism,
auxotrophies, etc...
On this basis, Bergey's Manual of Determinative Bacteriology was published in 1923, and its
subsequent editions became a reference in the field of bacterial taxonomy. Estimation of phylogeny
on the basis of phenotypic criteria (e.g. by a 'neighbour-matching' method) resulted in so-called
'phenenetic' classifications.

2
M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

These methods have made it possible to set up systematic approaches for the laboratory identification
and analysis of certain bacterial strains, including pathogens.

Figure: Example of the results of an API (analytical profile index) 20E gallery in which a strain of
Escherichia coli has been grown

So-called "chemotaxonomy" approaches based on the analysis of cellular components (lipids,

carbohydrates, proteins, specialized metabolites - localized in the wall, membranes, etc.) are also
described. Currently, some approaches based on mass spectrometry are being developed to analyze
cellular components in a global and rapid way for the identification of bacterial strains.
Finally, it should be noted that the classification of viruses is also partly based on morphological
criteria (presence or absence of an envelope, geometry of the capsid, number of capsomers, etc...).

Virus demarcation criteria:

- Family:
o Nature of the nucleic acid
o Viral symmetry
o Presence or absence of an envelop
- Groups and genus:
o Number of capsomers (if icosaedric)
o Nucleocapsid diameter (if
helicoidal)
o Origin of the envelop (if enveloped)
- Species: genomic sequence, natural host,
cell tropism, pathogenicity, mode of
transmission, physico-chemical properties,
antigenic properties

1.2. Limits
Phenotypic approaches (and phenotypic classification) nevertheless encounter important
limitations insofar as some traits are based on the functionality of genes susceptible to mutation (->
loss of the trait following inactivation) or on the acquisition by horizontal transfer of groups of genes
encoding characteristic functions (e.g. islands of pathogenicity). Thus E. coli and Shigella sp. are,
according to phenotypic criteria, the diseases they may (or may not cause for commensal strains of E.
coli) classified in different species when they are very similar from a genetic point of view. In addition,
monogenic polymorphisms and sexual dimorphism (e.g., in some fungi) can affect multiple
morphological traits.
Moreover, the number of characteristics that can be used as classification criteria remains
limited, so this approach is still of low power to account for microbial diversity both in terms of size
and precision (sensitivity, specificity). If convergent characteristics are used as classification criteria,
unrelated forms may be confused. In addition, approaches based on identification on certain media
apply only to crop organisms. Finally, these approaches may pose problems of reproducibility (see
Interpretation dependent on the experimenter).

3
M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

2. Species concept based on genome analysis

Molecular approaches therefore supplemented and then supplanted the phenotypic approach at
an early stage. The first method that emerged to define the species was based on the global
comparison of strain genomes by DNA-DNA hybridization (DDH) under standard conditions. Strains
were considered to be of the same species when the percentage of hybridization of their genomic DNA
was at least 70% (Wayne, Int J Syst Bacteriol 37:463-464, 1987). This technique made it possible for
the first time to classify microorganisms for which well-defined phenotypic characteristics were not
available.
This approach is a reference, a "gold standard" in bacterial taxonomy, but nevertheless has some
limitations, notably linked to the fact that it is based on a rather cumbersome (and error-prone)
protocol that must be carried out each time a new strain is cultivated (note the constraint linked to
the cultivable character of the organisms that can be studied in this way). The information collected
cannot be easily 'stored' on a computer and inferred from other data.

From O. Ogunseitan (2004)

This technique was later superseded by the ease of analysis of target gene sequences by PCR
followed by sequencing (initially by low-throughput sequencing techniques). At this level, the choice
of the reference gene is quite critical. It should in fact be a sort of "molecular clock" capable of
reflecting as accurately as possible the evolutionary distances between organisms. This concept
assumes a neutral and constant rate of evolution during evolution and within genomes, which raises
difficulties (see appendix on the concept of the molecular clock). For details on the molecular clock
concept see:
https://www.nature.com/scitable/knowledge/library/neutrality-and-molecular-clocks-
100492542/

4
M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

Very early on, 16S RNAs in prokaryotes (≈ 1500 bp), equivalent to 18S RNAs in eukaryotes,
genes universally present in cells, were used (Fox 1977, Woese 1987). It should be noted at this stage
that the classification of viruses poses additional problems in that no gene is universally present in
viruses. The analysis of the 16S RNA of archaea by Carl Woese will be the basis for the classification of
living beings in 3 domains: archaea, bacteria, eukaryotes. It should be noted that the question of the
organization of the tree of living beings with 2 (bacteria versus archaea-eukaryotes) or 3 branches
(bacteria, archaea, eukaryotes) is a debate that is experiencing a revival of interest since the discovery
of Lokiarchaea.
The question of the threshold of nucleic sequence identity at which two species should be
distinguished is highly critical and open to discussion. Currently, 16S RNA sequences with > 95%
identity are considered to represent the same genus, while sequences with > 97% identity represent
the same species. A threshold of 97% identity of the 16S RNA sequence is generally used to define an
operational taxonomic unit (OTU), which is widely used in microbial ecology and particularly useful
for assessing diversity within a sample containing non-cultivatable species. Although very practical,
this approach also has its limitations. For example, about half of the bacterial strains contain several
copies of 16S RNA (on average 4.2 - maximum 15), which are not always identical and may even diverge
by more than 97% (Vetrovsky & Baldrian, 2013). Furthermore, the variation in percent identity follows
a normal distribution that limits the effectiveness of threshold approaches.

Analysis based on 1690 published genomes - According to Vetrovsky & Baldrian (2013) - Rem: In this
paper, based on the 16S RNA sequence, Escherichia and Shigella were considered to belong to the
same genus.
Taking a single gene into account to classify living organisms is not a robust approach as it may
be subject to stochastic variation or acquisition by horizontal transfer. Molecular approaches based on
the analysis of several markers have therefore been developed. MLSA (multilocus sequence analysis)
or MLST (multilocus sequence typing), based on the analysis of the sequence of several so-called
"household" genes, has therefore been developed (Maiden, 1998). Here again, as with other
techniques developed subsequently, the definition of the threshold at which two species should be
considered as different posed difficulties. This approach is now entering the era of genomics, to take
into account the consideration of sequence identity between all the genes conserved between two
strains. In particular, the ANI (Average nucleotide identity test, Kostantinidis, 2006) is a method based
on the comparison of genomes and the calculation of the average identity of conserved genes. An ANI

5
M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

≥95% is classically used to define the species. It can be noted that this approach avoids the difficulties
linked to gene transfers by not taking them into account.

Analysis of 6787 prokaryotic genomes (normalization corresponds to the calculation of the average for
strains of the same species, some species being over-represented in terms of number of isolates) -
According to Kim et al. International Journal of Systematic and Evolutionary Microbiology, 2014
Viruses are classically considered to be different species if their sequence differs by more than
10%. However, viral diversity can be so great (due to genome replication using viral enzymes that are
not very faithful in some cases) that as early as 1993, Manfred Eigen proposed the notion of quasi-
species to designate a population (especially RNA viruses because of their greater variability) around
a "consensus sequence" (average) rather than diversity around a reference sequence.

The interest of approaches based on genome analysis is notably the possibility offered to study
non-cultivatable and/or uncultivated microorganisms by developing metagenomic approaches. These
have given rise to the concept of Co-Abundance Gene Groups (CAGs), which, when they contain more
than 700 genes, are referred to as metagenomic species (MGS) (Nielson et al., 2014), also known as
metagenome-assembled genomes (MAGs). These approaches can be used to characterize viral
genomes.

From Nielson et al., Nature Biotechnology, 2014

6
M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

3. Towards a species concept integrating the spatio-temporal and ecological dimensions of the
microbial world
3.1. Knowledge of the nucleic sequence is not enough...
Ecological aspects are likely to escape completely from approaches based solely on genome
analysis. In many cases, information on ecology (e.g. biotope, interactions) is often missing, but is still
key to maintaining a definition of the species centred on the notion of a population of individuals
sharing common characteristics.
However, these aspects are often at the heart of the concept of species. For example,
concerning viruses, the definition of a species according to Marc H. V. van Regenmortel in 1989 was
the following: the viral species is a polysthetic class (i.e. whose members share, in a non-absolute way,
several non-essential properties used to define the concept of species) constituting a continuous
lineage of replication (the various sequences follow one another, evolutionary lineage) possessing the
same ecological niche (in the broadest sense, i.e. in the same dimension of the relational properties of
viruses). The difficulty in grasping the concept of viral species led in 2008 to the adoption of a new, less
precise definition according to which the viral species is a group of related viruses occupying a
particular ecological niche, and whose members are so similar that it is better to have the same name
to designate them. The ecological dimension (and in particular the question of tropism) is therefore
within the notion of viral species, and is generally lacking in metagenomic databases. Approaches are
currently being developed to identify the host of certain prokaryotic viruses (CRIPSR sequence analysis,
EpicPCR for Emulsion, Paired Isolation and Concatenation PCR, metagenomic 3C-seq for chromosome
conformation capture).
However, this issue is at the heart of the debate concerning the inclusion of viral sequences
from metagenomic banks in the classification carried out by the International Committee on Virus
Taxonomy. For the moment, these are not taken into account, but some scientists believe that this
would lead to an underestimation of the number of viral species (by a factor of about 100 - see "The
consensus statement" by Simmons et al, Nature Reviews Microbiology, 2017).

3.2. Taking into account genome dynamics

Simpson's (1951) evolutionary conception of the species as a single lineage of ancestral
ancestry populations that retains its identity from other such lineages and has its own evolutionary
and historical trends is defeated in the case of the microbial species because evaluations of fully
sequenced microbial genomes (cellular or viral) have shown that many of them are in fact "hybrids".
For example, 5% to 65% of the genome of prokaryotic species can be acquired by horizontal transfer.
The acquisition of genes by horizontal transfer (typically conjugation, transduction, transformation in
prokaryotes, or even mechanisms involving nanotubes and vesicles) is likely to "muddy the waters" by
allowing the acquisition of new traits (phenotypic approaches) or sequences with a distinct
evolutionary history from the rest of the genome (genome-based approaches).
In this respect, certain processes are presented as being involved in a form of "species
barriers", likely to limit the entry of new genetic information into the cell, its insertion into the genome,
its maintenance and/or its expression, foreign sequences being likely to be toxic. Among these are :
the need for encounters between species to allow the exchange of material (not always possible if
different ecological niches, the need for physical contact during conjugation processes, etc.), the
mechanisms involved in the degradation of genomes (CRISPR system, DNA restriction-modification
systems), recombination frequency, exclusion mechanisms between mobile genetic elements,
mechanisms limiting the expression of sequences acquired by horizontal transfer (action of xenogeneic
silencers proteins, percentage in GC, CRISPR systems targeting RNA, RNAi, etc.). However, horizontal
transfers remain frequent. It is now known that this barrier is not robust, as gene transfers can take
place between different species, sometimes very far apart, sometimes between domains of living
organisms.
In order to reconcile the genomic species concept with ecological and evolutionary aspects,
Rossello-Mora and Amann (2001) proposed the "phylophenetic species concept" in which species are
defined as "a monophyletic and genomically coherent cluster of individual organisms that show a high

7
M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

degree of overall similarity in many independent characteristics, and is diagnosable by a discriminative

phenotypic property" (the species is "a monophyletically and genomically coherent group of
individuals that exhibit a high degree of general similarity in many independent characteristics and is
diagnosable by a discriminative phenotypic property"). This concept finds natural extension in
phylogenomic approaches to reconstruct the evolutionary history of organisms by considering whole
genomes or large fractions of genomes. These approaches are based on the construction of complete
sets of phylogenetic trees taking into account all the proteins of an organism. When these approaches
include proteins encoded by genes acquired by horizontal transfer, they make it possible to take into
account the entire evolutionary history that has led to the groups of individuals known as "species".
This consideration of genome dynamics leads to the definition, for a given species, of the
pangenomes (set of genes present in at least one representative of the species) and the core genome
(set of genes common to all the representatives of the species). The difference between the two
corresponds to the so-called "accessory" genome, which confers specificity to certain strains (e.g.
strains of uropathogenic E. coli). The size of these groups of genes seems to vary greatly from one
species to another, particularly in relation to its lifestyle. For example, less than 40% of the pangenome
genes belong to the core genome in Escherichia coli, while the intracellular pathogen Chamydia
trachomatis has a pangenome barely larger than the core genome (974 versus 821 genes,
respectively).

8
M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

From Brockhurst et al., Current Biology (2019)

3.3 Thinking on the scale of the ecotype

If we take the definition of species according to which "a species is a group of organisms whose
divergence is limited by a cohesive force; divergence between different species is irreversible; and
different species are ecologically distinct", the properties described are not held by the concept of
microbial species as described so far but by the concept of ecotype as stated in 2002 by Frederik Cohan.
According to Cohan, ecotypes as "populations of organisms occupying the same ecological niche,
whose divergence is purged recurrently by natural selection". This definition of the prokaryotic
ecotype, situated in terms of scale below the definition of the prokaryotic species, is however in its
essence closer to the notion of "biological" species as defined in eukaryotes. It takes into account the
evolutionary history and ecological adaptations common to the individuals of an ecosystem. Although
different in modalities, periodic selection and the formation of barriers to gene flow between ecotypic
species echoes the notion of species in sexed organisms.
Ecotypes are subject to diversification and selection, which can lead to speciation (formation
of a new species).

9
M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

From F. Cohan, Microbiol. Spectrum, 2017

10
M2 Fundamental Microbiology – Pr Stéphanie Bury-Moné – University of Paris-Saclay

Take home messages:

- The evolving concept of species in microbiology is currently being grasped:
- with the help of phenotypic tests in particular within the framework of the analysis of certain
samples in analysis laboratories.
- by molecular approaches (also available in analytical laboratories) or even genomics or
metagenomics to analyze microbial diversity on complex samples
However, this concept of species sometimes remains insufficiently precise to account for the specificity
of certain interactions within microbial ecosystems, which led to the development of the concept of
ecotype.
- Reflection on the question of microbial species has highlighted the importance of considering the
dynamics of genomes on time scales that can be quite short (in relation to the rapid proliferation of
certain microorganisms). While certain processes contribute to ensuring a certain stability of genomes
(e.g. mechanisms of DNA replication and repair, barriers against the acquisition or maintenance of
foreign regions), these remain subject to variation, which contributes to the evolution of genomes. We
can therefore only capture a picture of the diversity of living things at a given time, knowing that it is
constantly changing.
- The field of synthetic biology has been opening up over the last few decades, leading to the
development of genetically modified organisms through genetic engineering and its evolution in terms
of synthetic biology. This raises the question of the position of so-called "synthetic" species in this tree
of life. => Moreno E (2012) Design and Construction of “Synthetic Species”. PLoS ONE 7(7): e39054.
doi:10.1371/journal.pone.0039054

Also see: “Microbial Diversity: Form and Function in Prokaryotes” (2004) de Oladele Ogunseitan :
http://www.blackwellpublishing.com/content/BPL_Images/Content_store/Sample_chapter/063
2047089/Ogunseitan_sample%20chapter_Microbial%20Diversity.pdf

11