Bacterial Stigmergy: An Organising Principle of Multicellular Collective Behaviours of Bacteria

Abstract

The self-organisation of collective behaviours often manifests as dramatic patterns of emergent large-scale order. This is true for relatively "simple" entities such as microbial communities and robot "swarms," through to more complex self-organised systems such as those displayed by social insects, migrating herds, and many human activities. The principle of stigmergy describes those self-organised phenomena that emerge as a consequence of indirect communication between individuals of the group through the generation of persistent cues in the environment. Interestingly, despite numerous examples of multicellular behaviours of bacteria, the principle of stigmergy has yet to become an accepted theoretical framework that describes how bacterial collectives self-organise. Here we review some examples of multicellular bacterial behaviours in the context of stigmergy with the aim of bringing this powerful and elegant self-organisation principle to the attention of the microbial research community.

Full text

Review Article
Bacterial Stigmergy: An Organising Principle of
Multicellular Collective Behaviours of Bacteria
Erin S. Gloag, Lynne Turnbull, and Cynthia B. Whitchurch
e ithree Institute, University of Technology Sydney, P.O. Box 123, Broadway, Sydney, NSW 2007, Australia
Correspondence should be addressed to Cynthia B. Whitchurch; cynthia.whitchurch@uts.edu.au
Received  August ; Revised  December ; Accepted  December 
Academic Editor: Pascal Vallotton
Copyright ©  Erin S. Gloag et al. is is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
e self-organisation of collective behaviours oen manifests as dramatic patterns of emergent large-scale order. is is true for
relatively “simple” entities such as microbial communities and robot “swarms,” through to more complex self-organised systems
such as those displayed by social insects, migrating herds, and many human activities. e principle of stigmergy describes those
self-organised phenomena that emerge as a consequence of indirect communication between individuals of the group through the
generation of persistent cues in the environment. Interestingly, despite numerous examples of multicellular behaviours of bacteria,
the principle of stigmergy has yet to become an accepted theoretical framework that describes how bacterial collectives self-organise.
Here we review some examples of multicellular bacterial behaviours in the context of stigmergy with the aim of bringing this
powerful and elegant self-organisation principle to the attention of the microbial research community.
1. Introduction
e emergence of self-organised patterns in living and non-
living systems has fascinated scientists for centuries. In bio-
logical systems, the coordination of group behaviours and the
subsequent emergence of large-scale patterns are inherently
more complex than that which spontaneously emerges in
nonliving systems [], involving an interplay of physical,
chemical, and biological interactions, both physiological and
behavioural, that have been honed through natural selection
[–]. Many self-organised phenomena in both biotic and
abiotic systems can be explained by the principle of stig-
mergy, a concept that describes self-organised systems that
arisethroughanindividualofthecollectiveinuencingthe
movement or behaviour of other individuals at a later point
in time through the generation of persistent cues within the
environment [,].
e concept of stigmergy was rst introduced by the
entomologist Grass´
e in  to explain the construction of
termite colonies []. is powerful concept, for the rst time,
explained how apparently random and independent move-
ments of an individual could result in the transfer of persis-
tent information locally, thereby manifesting as coordinated
behaviour at a global level [,]. e principle of stigmergy
has since been employed to describe a vast array of group
activities such as the laying-down of pheromone trails by
foraging ants, herd migration in animals, and various aspects
of human activities including the following of hiking trails
and pedestrian footpaths [–] as well as articial systems
such as “swarm intelligence” within robotics and computing
[–]. Interestingly, even the development of multicellular
tissues has been described as a stigmergic phenomenon in
which chemical cues are embedded in extracellular matrix
material []. As other scientic elds such as social sci-
ences, technology, and computer sciences began adopting the
concept of stigmergy to help describe and explain various
phenomena of emergent behaviour or properties, various
types or categories of stigmergy have been described in
an attempt to better understand the dierent stimuli and
responses which inuence the stigmergic interactions of the
agents in these systems.
Sematectonic and marker-based stigmergy dierentiate
between the forms of communication, that is, the types of
signals that initiate a response or behaviour change [,,
]. Sematectonic stigmergy was rst coined by Wilson and
describes communication through physical changes to the
Hindawi Publishing Corporation
Scientifica
Volume 2015, Article ID 387342, 8 pages
http://dx.doi.org/10.1155/2015/387342

Scientica
environment [], for example, the following of trails by
herd animals, pedestrians, and hikers []andtheconstruc-
tion of wasp nests, where the development of the physical
structure acts as cues for the next steps or process in the
design [,]. In contrast, marker-based stigmergy refers to
communication through the deposition of chemical signals
within the environment [,,], for example, the following
of pheromone trails by ants aiding in their food foraging
behaviour []. A further distinction between these two vari-
ations is that for sematectonic stigmergy the communicative
information tends to provide a direct contribution to the
task/emergent property, whereas in marker-based stigmergy
the cues do not take direct action but rather inuence
subsequent behaviour, stimulating self-organisation for its
eective completion [].
Quantitative and qualitative stigmergy are other forms
that were introduced by eraulaz and Bonabeau to describe
the dierent stimulus signalling and response outcomes [].
Quantitative stigmergy describes a system where an individ-
ual’s response to a stimulus intensies the stimulus, with the
nature of these stigmergic systems oen leading to positive
feedback [,]. Here again the following of pheromone
trails by ants provides an example of quantitative stigmergy,
whereby continuous ant trac along specic pheromone
trails results in the deposition of more pheromone, thereby
amplifying the signal and attracting further ants to these
trails,whichinturnlaydownmorepheromone.Qualitative
stigmergy refers to self-organising systems that arise from
an individual responding to a stimulus that in turn creates a
qualitatively dierent stimulus, thereby triggering a separate
response [,]. e building of a wasp nest provides an exam-
ple of qualitative stigmergy as the growing structure provides
dierent signals and cues based on the stage of construction
and results in distinct building behaviours [,,]. It has
also been recognised that both sematectonic and marker-
based signals can initiate either quantitative or qualitative
responses []. Finally, passive and active stigmergy have
been described; however, as these two variations have for the
most part been applied only to collective swarm intelligence
in robotics [,], they will not be discussed here.
Whilst there are many examples of self-organised mul-
ticellular behaviours of bacteria, the concept of stigmergy
has rarely been used to describe these phenomena. Here we
review some examples of bacterial collective behaviours that
may be described in the context of the organising principle of
stigmergy.
2. Bacterial Swarms
Many species of bacteria are able to actively migrate across
surfaces via a number of dierent mechanisms including
twitching, gliding, and agella-mediated swarming motili-
ties. ese motilities can facilitate the surface translocation
of individual cells but can also manifest as highly organised
multicellular “swarms” that enable rapid expansion of the
bacterial communities. Here we show that stigmergy explains
many of these “swarming” behaviours of bacteria.
Twitching motility is a mechanism of surface transloca-
tion that is powered by the extension, surface binding, and
subsequent retraction of type IV pili (tfp) [,]. Under
appropriate conditions, twitching motility is as a complex,
highly coordinated multicellular behaviour that leads to the
active expansion of the bacterial community across solidied
nutrient media [–]. It has been found that when Pseu-
domonas aeruginosa is cultured at the interface of nutrient
media that has been solidied with .% gellan gum and
an abiotic surface such as plastic or glass, twitching motility
can lead to the formation of highly structured multicellular
communities at the interstitial space. ese have several char-
acteristic micromorphological features including large van-
guard ras of highly aligned cells at the leading edge behind
which there is an intricate, interconnected lattice-like net-
work of trails of cells (Figure (a);[]). Semmler et al. pro-
posed that as the vanguard ras migrated across the surface
of the semisolid nutrient media, they created some form of
trail along which ensuing cells preferentially followed [].
We have shown recently that the emergence of the inter-
connected network of trails is likely to occur due to the for-
mation of an interconnected furrow system in the underlying
semisolid media (Figure (b);[,]). We recently applied
the concept of stigmergy to describe the emergent self-
organisation of P. a e r u g i n o s a interstitial communities that
occurs as a consequence of cells creating and travelling within
the furrow network []. To our knowledge this was the rst
description of stigmergic behaviour driven by physical cues
within the environment. We hypothesised that as cellular
aggregates migrated across the media surface, they forged
a furrow along which ensuing cells migrated and in doing
so continued to remodel the substratum thereby rening
the furrow system [,]. We proposed that the furrows
physically conne cells thereby directing cell movement and
contributing to the emergence of the intricate interconnected
network of cellular trails that are a characteristic feature
of these biolms (Figure (a);[,]). is is an example
of sematectonic and quantitative stigmergy and is highly
reminiscent of the physical trail following displayed by
animals during herd migrations and by humans following
hiking trails and pedestrian footpaths [–,].
Interestingly, some bacteria from diverse genera display
an “agar pitting” phenotype which can be used as an identi-
fying feature for these species [–]. One such organism,
Dichelobacter nodosus, also produces striking interconnected
pattern networks reminiscent of that of P. a e r u g i n o s a when
they are grown at the interstitial space between the petri
dish and media []. However, whether this emergent pattern
arises due to the corrosion of the agar during biolm
expansion, creating furrows that guide the movements of
the bacteria remains to be determined. e agar pitting
phenotypes of both D. nodosus and Moraxella bovis have been
correlated with the presence of tfp. It has been speculated that
the agar polysaccharides may act as ligands to which the tfp
bind and that the physical interaction of the tfp with the agar
may be responsible for the agar pitting phenotype []. It is
interesting to speculate that the formation of furrow networks
may constitute a more global mechanism for the stigmergic
organisation of bacterial communities.
We have recently also identied a role for extracellular
DNA (eDNA) in coordinating the collective behaviour of

Scientica 
(a)
0 100 200
(nm)
200.0
100.0
0.0
(nm)
60.0
45.0
30.0
15.0
0.0 0.0
0.0
15.0
30.0
45.0
60.0
(𝜇m)
(𝜇m)
(b) (c)
(d) (e)
F : Stigmergic self-organisation of bacterial communities. (a) Pseudomonas aeruginosa interstitial biolm imaged using phase contrast
microscopy depicting the emergent pattern formation. At the advancing edge are ras of cells that initiate biolm expansion, behind which
there is an interconnected lattice-like network of cellular trails. Scale bar indicates  𝜇m. (b) D rendered image of the interconnected furrow
network underlying the P. a e r u g i n o s a interstitial biolms imaged using atomic force microscopy(AFM) within the lattice-like network. Height
scale is relative. (c) P. a e r u g i n o s a expressing cyan uorescent protein (CFP; blue) interstitial biolms were grown on media supplemented
with the cell impermeant nucleic acid dye TOTO- to visualize eDNA (yellow) and imaged using OMX BLAZE wide-eld microscopy. Scale
bar indicates  𝜇m. Swarming communities of (d) Pr. vulgaris and (e) M. xanthus grown on semisolid nutrient media and imaged using phase
contrast microscopy revealing the phase bright trails routinely observed at the leading edge. Scale bar is  𝜇m.

Scientica
P. a e r u g i n o s a cells undergoing twitching motility-mediated
biolm expansion []. We observed that P. a e r u g i n o s a
interstitial biolms contain eDNA distributed either as a
ne coating of the cells or as concentrated, punctate foci
from which thin tendrils radiated in the overall direction
of the motion of cells (Figure (c);[]). Removal of this
eDNA, through the incorporation of the eDNA-degrading
enzyme DNaseI into the solidied nutrient media, resulted
in the abrogation of the characteristic interconnected pattern
network of these biolms [].
To understand the role of eDNA within these biolms
we employed a computer vision and cell tracking analysis
pipeline [,,]toquantitatethebehaviourofthe
individual cells in the absence and presence of DNaseI.
Interrogation of the resulting image informatics database
revealed that eDNA facilitates twitching motility-mediated
biolm expansion by enabling more frequent movements of
individual cells, thereby resulting in more sustained motion
and greater distances traversed by individual cells over longer
time periods. ese analyses also revealed that eDNA is
required for maintaining coherent cell behaviour and cell
alignment over time []. Previous reports have identied
that P. a e r u g i n o s a tf p bind to DNA []andthatP. a e r u g i n o s a
cells spontaneously pneumatically orient with the direction
of extended DNA chains in a matrix of aligned, concentrated
DNA []. We proposed that the bed of aligned eDNA
molecules within P. a e r u g i n o s a interstitial biolms maintains
cell orientations by aligning cells to the thin strands of
eDNA and that eDNA provides a substrate for optimal tfp
binding, consequently enabling more frequent tfp-powered
translocations, ensuring smooth trac ow within the trail
network and a consistent supply of cells to the leading
edge of the expanding biolm []. We also propose that
eDNA serves as an intercellular “glue” that binds the cells
together within vanguard ra assemblages thereby facilitating
coherent cell movements to power migration of leading edge
ras into virgin territory [].
e ability of eDNA to promote cohesive group behaviour
during active biolm expansion is an example of sematec-
tonic stimergy. It could be further argued that the redistribu-
tion of eDNA through the biolm is also an example of quan-
titative stigmergy as continued cellular migration through the
concentrated regions of eDNA results in the production of
ne tendrils of eDNA aligned with the direction of bacterial
migration which then directs and maintains the alignment of
ensuing cells along these eDNA strands thereby maintaining
trac ow in the overall direction of travel of the preceding
cells [].
Flagella-mediated swarming motility of Proteus spp. leads
to the formation of rapidly expanding colonies grown on agar
that are characterised by a repeated concentric circle pattern
that extends across the swarm. is patterning is attributed to
continuous rounds of cell dierentiation, where the normal
rod cells, which are largely nonmotile, dierentiate into long,
hyperagellated swarmer cells. As a collective these swarmer
cells rapidly migrate across the surface until they dierentiate
back to the nonmotile normal cells resulting in consolidation
and the formation of the observed ring pattern [,]. e
agella of Proteus swarmer cells interweave with agella from
the same cell and with those of neighbouring cells, forming a
connected and highly synchronised swarming front that aids
in the rapid expansion by these colonies []. e secretion
of an extracellular slime has been found to facilitate the
collective swarming behaviour of Pr. mirabilis. At the leading
edge of Pr. mirabilis swarms, swarmer cells are encased in
a slime layer and appear to preferentially move along an
interconnected network of phase bright slime trails (Fig-
ure (d);[]). It has been hypothesised that the slime trails
aid in directing swarming motility and the slime encasement
facilitates the maintenance of a cohesive organisation of cells
[,]. erefore slime production and slime trail follow-
ing promote the self-organisation of collective behaviours
necessary for the expansion of the swarming colony.
Gliding motility of Myxococcus xanthus is mediated by
the combined eorts of two motility modes; social (S)
motility and adventurous (A) motility. Similar to twitching
motility, S-motility is driven by the extension, binding, and
retraction of tfp with this motility mode being typically
displayed by groups or clusters of cells [,]. A-motility
mediates single cell migration and in contrast to that of
S-motility, the machinery driving A-motility is yet to be
conrmed and is an area of controversy [,]. However
all current schools of thought predict the role of a secreted
slime in facilitating the A-motility of this organism [–
], where phase bright trails are observed at the leading
edge of the M. xanthus swarms when grown on semisolid
media (Figure (e);[]), similar to that of Pr. mirabilis.
M. xanthus cells preferentially migrate along these slime
trails, with cells frequently observed to turn onto the trails
rather than migrating across virgin territory. Continued
cellular trac along the trails results in their thickening and
extension as a consequence of continued slime deposition
[,]. It is recognised that this trail following behaviour
coordinates the collective behaviour of M. xanthus cells,
specically those displaying A-motility, at the leading edge of
the surface swarms, and contributes to the emergence of the
interconnected pattern networks at these areas [–].
e following of slime trails during Pr. mirabilis swarming
and M. xanthus gliding motilities are both sematectonic
and quantitative stigmergic systems, where the stimulus
(slime) is a physical manifestation within the environment
that directly contributes to the expansion of the community
as it is required for the motility of the organism. is is
particularly the case for the slime mediating the A-motility
of M. xanthus (sematectonic stigmergy). Continued trac
along the slime trails amplies the slime deposited resulting
in further recruitment of cells migrating along these regions
(quantitative stigmergy).
It has been shown recently that the formation of vor-
texes comprised of thousands of bacteria rotating in unison
that occur during active surface migration by Paenibacillus
vortex biolms occurs as a consequence of the actions of a
subpopulation of lamentous cells that direct the motion of
the other members of the collective [,]. is appears to
be another example of bacterial stigmergy, though it remains
to be determined if this collective behaviour occurs as a
consequence of physical alteration of the environment, slime,
or chemical cues.

Scientica 
A number of computational models have been devel-
oped to describe collective behaviours displayed during
swarm activities, particularly for M. xanthus [–]. Due
to the inherent diculties in modelling biological systems,
a number of these models do not truly reect experimental
observations or contain artefacts as a consequence of the
rule parameters incorporated into the model []. Stigmergic
systems have long been the focus of extensive computational
modelling to understand the emergent properties within
these systems [–] and to relate stigmergic principles from
one system to another in an attempt to draw comparisons
from well-studied and established systems []. It is our con-
tentionthatasimilarapproachcouldbetakenformodelling
bacterial swarming communities through the incorporation
of key ideas from other stigmergic models, such as those
ofHelbingetal.andGoldstoneetal.,whomodelledsema-
tectonic and quantitative stigmergic systems such as trail
following by humans and animals [–]. Incorporating such
an integrated approach could potentially yield further insight
into the self-organisation and emergent pattern networks of
bacterial swarms.
3. Bacterial Biofilms
Bacterial biolms are multicellular communities of bacteria
thatareattachedtoeachotherandoenabioticorabi-
otic surface via a self-produced extracellular matrix com-
prised of extracellular polymeric substances (EPS) including
exopolysaccharides, eDNA, proteins, and lipids [–]. e
production of this EPS matrix is essential for biolm develop-
ment as it provides intercellular connectivity that binds cells
to each other and, in the case of surface-attached biolms,
provides surface adherence [,]. e ability of the EPS
matrix produced by biolm cells to promote cohesion and
surface attachment of the biolm community is an example
of sematectonic stigmergy.
It has been observed that individual P. a e r u g i n o s a
cells undergo extensive twitching motility-mediated surface
exploration prior to subsequent microcolony formation dur-
ing the early stages of biolm formation on glass submerged
in liquid nutrient media [–]. Zhao and colleagues
showed recently that, during surface exploration, P. a e r u g i -
nosa cells deposited trails of the exopolysaccharide Psl, which
appeared to recruit additional cells along these trails leading
to a positive feedback loop of further Psl deposition and
subsequent cell attraction []. It was hypothesised that
this trail following behaviour was facilitated by twitching-
motility-mediated surface exploration, where the tfp were
thought to probe the surrounding areas for Psl networks,
promoting binding of the tfp and directing cellular migration
to these areas []. In areas of high Psl concentration, cells
were observed to adhere to the substratum and correlated to
the subsequent sites of microcolony formation [,]. is
mechanism of following exopolysaccharide trails to coor-
dinate the single cellular motilities of P. a e r u g i n o s a during
early biolm development is an example of sematectonic
and quantitative stigmergy. Zhao et al. used a “rich-getting-
richer” analogy comparable to that of capitalist economies to
describe this emergent behaviour [], which has itself been
described as a stigmergic system [,].
4. Quorum Sensing
In many bacterial communities quorum sensing regulates
and coordinates social behaviours, such as bioluminescence,
secretion of public goods, and the switch from planktonic
to the biolm mode of growth []. Quorum sensing
occurs through the release of small molecules by individual
bacteria into the environment by passive diusion. e
concentration of these small molecules increases within the
environment with increasing cell density, permitting cells
to gather information about their surrounding neighbours.
Once a sucient quantity of signal is present within the
environment, reecting a critical population density, a gene
regulation cascade is initiated culminating in the up- or
downregulation of the expression of various genes required
for social behaviours, virulence factor production, and so
forth [,]. In this manner it has been identied that
quorum sensing can regulate the expression of over  genes
within P. a e r u g i n o s a [].
Quorum sensing within bacterial communities bares a
striking resemblance to pheromone signalling that coordi-
nates the collective behaviours of social insects. It is there-
fore interesting to speculate whether quorum sensing oers
another example of stigmergic self-organisation within bacte-
rial communities. Under circumstances where quorum sens-
ing signalling molecules are able to persist and accumulate
withintheenvironment,thenquorumsensingcouldbe
considered an example of marker-based stigmergy whereby
the release of signalling molecules into the environment
stimulates collective behaviours of the growing bacterial
population, similar to the pheromones coordinating the
social behaviours of ants and termites. It could be suggested
that quorum sensing, in addition to marker-based stigmergy,
is also an example of qualitative stigmergy, where, depending
on their concentration, the quorum sensing signals trigger
dierent responses by the bacterial population.
5. Summary and Future Directions
We have presented a number of examples in which bacteria
employ stigmergic self-organisation to coordinate their col-
lective behaviours and found that sematectonic and quanti-
tative stigmergic systems in the form of trail following were
the most prevalent in the above examples. is highlights
the conserved nature of self-organising mechanisms within
nature regardless of the cognitive abilities of the individual
entities and suggests a common evolution of trail following
as a simple yet eective means of coordinating collective
behaviours.
e idea that self-organising systems utilised by bacterial
communities are similar to those utilised by higher organisms
is gaining interest within the scientic community. A recent
reviewhascalledfortheemploymentofamoreintegrative
approach across scientic elds in the study of self-organising
systems []. Stigmergy provides an excellent example of
this approach where, since its rst introduction within the
eld of entomology [], the importance of this concept has
been recognised across diverse areas ranging from biology
to social sciences, technology, and computer sciences [,
,]. e wide acceptance of stigmergy can, for the most

Scientica
part, be attributed to a special edition of Articial Life
dedicated to stigmergic systems [,], with the hopes of
bringing this concept to the forefront within the scientic
community. is concept, despite its obvious application to
the understanding of multicellular bacterial behaviours, has
been largely overlooked within the eld of microbiology.
Here we recognise the importance of the concept of stig-
mergic self-organisation and the implications it has on under-
standingthecollectivebehavioursofcomplexmulticellular
bacterial communities. We propose that bacterial stigmergy
should be included in the repertoire of systems that bacteria
employ to control multicellular activities. Furthermore, we
suggest that bacterial stigmergic systems may provide testable
models to explore stigmergic self-organisation at a molec-
ular level [], which is currently an unexplored concept
and will ultimately lead to greater understanding of other
biological stigmergic systems. Understanding the mecha-
nisms employed by bacteria to coordinate their multicellular
behaviours may lead to the development of novel strategies
to control infections and biofouling in industrial and marine
settings.
Conflict of Interests
e authors declare that there is no conict of interests
regarding the publication of this paper.
Acknowledgment
Cynthia B. Whitchurch was supported by a NHMRC Senior
Research Fellowship ().
References
[] V. Narayan, N. Menon, and S. Ramaswamy, “Nonequilibrium
steady states in a vibrated-rod monolayer: tetratic, nematic, and
smectic correlations,” Journal of Statistical Mechanics: eory
and Experiment,no.,ArticleIDP,.
[] S. Camazine, J. L. Deneubourg, N. Franks, J. Sneyd, G. er-
aulaz, and E. Bonabeau, Self-Organization in Biological Systems,
Princeton University Press, Princeton, NJ, USA, .
[] B.Grammaticos,M.Badoual,andM.Aubert,“An(almost)solv-
able model for bacterial pattern formation,” Physica D: Nonlin-
ear Phenomena,vol.,no.,pp.–,.
[] H. Levine and E. Ben-Jacob, “Physical schemata underlying
biological pattern formation—examples, issues and strategies,”
Physical Biology,vol.,no.,pp.P–P,.
[] P.-P. Grass´
e, “La reconstruction du nid et les coordinations
interindividuelles chez Bellicositermes natalensis et Cubitermes
sp. la th´
eorie de la stigmergie: Essai d’interpr´
etation du com-
portement des termites constructeurs,” Insec tes Soc iaux,vol.,
no. , pp. –, .
[] G. eraulaz and E. Bonabeau, “A brief history of stigmergy,”
Articial Life,vol.,no.,pp.–,.
[] R.L.GoldstoneandM.E.Roberts,“Self-organizedtrailsystems
in groups of humans,” Complexity,vol.,no.,pp.–,.
[] D.Helbing,J.Keltsch,andP.Moln
´
ar, “Modelling the evolution
of human trail systems,” Nature,vol.,no.,pp.–,
.
[]D.Helbing,F.Schweitzer,J.Keltsch,andP.Moln
´
ar, “Active
walker model for the formation of human and animal trail
systems,” Physical Review E,vol.,no.,pp.–,.
[] E.Boissard,P.Degond,andS.Motsch,“Trailformationbased
on directed pheromone deposition,” Journal of Mathematical
Biology,vol.,no.,pp.–,.
[] E. Bonabeau, G. eraulaz, J.-L. Deneubourg, S. Aron, and
S. Camazine, “Self-organization in social insects,” Trend s in
Ecology and Evolution,vol.,no.,pp.–,.
[] O. Holland and C. Melhuish, “Stigmergy, self-organization, and
sorting in collective robotics,” Articial Life,vol.,no.,pp.
–, .
[] J.Stradner,R.enius,P.Zahadat,H.Hamann,K.Crailsheim,
and T. Schmickl, “Algorithmic requirements for swarm intelli-
gence in dierently coupled collective systems,” Chaos, Solitons
&Fractals,vol.,pp.–,.
[] L. R. Christensen, “Stigmergy in human practice: coordination
in construction work,” Cognitive Systems Research,vol.,pp.
–, .
[] L. Marsh and C. Onof, “Stigmergic epistemology, stigmergic
cognition,” Cognitive Systems Research,vol.,no.-,pp.–
, .
[] H. V. D. Parunak, “A survey of environments and mechanisms
for human-human stigmergy,” in Environments for Multi-Agent
Systems II,vol.ofLecture Notes in Computer Science,pp.
–, Springer, Berlin, Germany, .
[] P. Peeters, H. van Brussel, P. Valckenaers et al., “Pheromone
based emergent shop oor control system for exible ow
shops,” Articial Intelligence in Engineering,vol.,no.,pp.
–, .
[] S. Burbeck, “Complexity and the evolution of computing: bio-
logical principles for managing evolving systems,” Computing
Systems, pp. –, .
[] E. O. Wilson, Sociobiology: e New Synthesis,HarvardUniver-
sity Press, Cambridge, Mass, USA, .
[] H. A. Downing and R. L. Jeanne, “Nest construction by
the paper wasp, Polistes: a test of stigmergy theory,” Animal
Behaviour, vol. , no. , pp. –, .
[] S.Goss,R.Beckers,J.-L.Deneubourg,S.Aron,J.Pasteels,and
R. N. Hughes, “How trail laying and trail following can solve
foraging problems for ant colonies,” in Behavioural Mechanisms
of Food Selection, pp. –, .
[] O. Holland, “Multi-agent systems: lessons from social insects
and collective robotics,” in Proceedings of the AAAI Spring Sym-
posium on Adaptation, Coevolution and Learning in Multiagent
Systems, pp. –, AAAI Press, Menlo Park, Calif, USA, .
[] A. J. Merz, M. So, and M. P. Sheetz, “Pilus retraction powers
bacterial twitching motility,” Nature, vol. , no. , pp. –
, .
[] J. M. Skerker and H. C. Berg, “Direct observation of extension
and retraction of type IV pili,” Proceedings of the National
Academy of Sciences of the United States of America, vol. , no.
, pp. –, .
[] A. B. T. Semmler, C. B. Whitchurch, and J. S. Mattick, “A re-
examination of twitching motility in Pseudomonas aeruginosa,”
Microbiology,vol.,no.,pp.–,.
[] K. Bovre and L. O. Froholm, “Variation of colony morphology
reecting mbriation in Moraxella bovis and two reference
strains of M. nonliquefaciens,” Acta Pathologica et Microbiolog-
ica Scandinavica—Section B: Microbiology and Immunology,vol.
,no.,pp.–,.

Scientica 
[] J. C. McMichael, “Bacterial dierentiation within Moraxella
bovis colonies growing at the interface of the agar medium with
the Petri dish,” Journal of General Microbiology,vol.,no.,
pp. –, .
[] E. S. Gloag, L. Turnbull, A. Huang et al., “Self-organization of
bacterial biolms is facilitated by extracellular DNA,” Proceed-
ings of the National Academy of Sciences of the United States of
America, vol. , no. , pp. –, .
[] E. S. Gloag, M. A. Javed, H. Wang et al., “Stigmergy: a key driver
of self-organization in bacterial biolms,” Communicative and
Integrative Biology,vol.,no.,ArticleIDe,.
[] J. J. Alexander and J. F. Lew is, “Pitting of agar surfac e by
Pseudomonas stutzeri,” Journal of Clinical Microbiology,vol.,
article , .
[] S. D. Henriksen, ““Pitting” and “corrosion” of the surface of
agar cultures by colonies of some bacteria from the respiratory
tract,” Acta Pathologica Microbiologica Scandinavica Section B
Microbiology and Immunology,vol.,no.,pp.–,.
[] J. M. Tennent and J. S. Mattick, “Type  mbriae,” in Fimbriae
Adhesion, Genetics, Biogenesis, and Vaccines, P. Klemm, Ed., pp.
–, CRC Press, .
[]S.H.Zinner,A.K.Daly,andW.M.McCormack,“Isolation
of Eikenella corrodens in a general hospital,” Journal of Applied
Microbiology,vol.,no.,pp.–,.
[] D.N.Love,R.F.Jones,M.Bailey,andA.Calverley,“Comparison
of strains of gram-negative, anaerobic, agar-corroding rods
isolated from so tissue infections in cats and dogs with
type strains of Bacteroides gracilis,Wol i n e l l a recta,Wolinel l a
succinogenes,andCampylobacter concisus,” Journal of Clinical
Microbiology,vol.,no.,pp.–,.
[] J.V.A.RobinsonandA.L.James,“Someobservationsonthe
colony morphology of “corroding bacilli”,” Journal of Applied
Bacteriology,vol.,no.,pp.–,.
[] X. Han, R. M. Kennan, J. K. Davies et al., “Twitching motility
is essential for virulence in Dichelobacter nodosus,” Journal of
Bacteriology,vol.,no.,pp.–,.
[] P. Vallotton, L. Mililli, L. Turnbull, and C. Whitchurch, “Seg-
mentation of dense D bacilli populations,” in Digital Image
Computing: Techniques and Applications,DICTA,Sydney,Aus-
tralia, .
[] P. Vallotton, C. Sun, D. Wang, L. Turnbull, C. Whitchurch,
and P. Ranganathan, “Segmentation and tracking individual
pseudomonas aeruginosa bacteria in dense populations of
motile cells,” in Proceedings of the 24th International Conference
Image and Vision Computing New Zealand (IVCNZ ’09),pp.
–, Wellington, New Zealand, November .
[] E. J. van Schaik, C. L. Giltner, G. F. Audette et al., “DNA
binding: a novel function of Pseudomonas aeruginosa type IV
pili,” Journal of Bacteriolog y,vol.,no.,pp.–,.
[] I.I.Smalyukh,J.Butler,J.D.Shrout,M.R.Parsek,andG.C.L.
Wong, “Elasticity-mediated nematiclike bacterial organization
in model extracellular DNA matrix,” Physical Review E—
Statistical, Nonlinear, and So Matter Physics,vol.,no.,
Article ID , .
[] O. Rauprich, M. Matsushita, C. J. Weijer, F. Siegert, S. E. Esipov,
andJ.A.Shapiro,“PeriodicphenomenainProteus mirabilis
swarm colony development,” Journal of Bacteriolog y,vol.,
no.,pp.–,.
[] R. M. Harshey, “Bees aren’t the only ones: swarming in gram-
negative bacteria,” Molecular Microbiology,vol.,no.,pp.
–, .
[] B. V. Jones, R. Young, E. Mahenthiralingam, and D. J. Stickler,
“Ultrastructure of Proteus mirabilis swarmer cell ras and role
of swarming in catheter-associated urinary tract infection,”
Infection and Immunity,vol.,no.,pp.–,.
[] S. J. Stahl, K. R. Stewart, and F. D. Williams, “Extracellular slime
associated with Proteus mirabilis during swarming,” Journal of
Bacteriology,vol.,no.,pp.–,.
[] J. Hodgkin and D. Kaiser, “Genetics of gliding motility in Myx-
ococcus xanthus (Myxobacterales): two gene systems control
movement,” Molecular and General Genetics,vol.,no.,pp.
–, .
[] D. Kaiser, “Social gliding is correlated with the presence of pili
in Myxococcus xanthus,” Proceedings of the National Academy of
Sciences of the United States of America, vol. , no. , pp. –
, .
[] E. M. F. Mauriello, T. Mignot, Z. Yang, and D. R. Zusman,
“Gliding motility revisited: how do the myxobacteria move
without agella?” Microbiology and Molecular Biology Reviews,
vol.,no.,pp.–,.
[] T.Mignot,“eelusiveengineinMyxococcus xanthus gliding
motility,” Cellular and Molecular Life Sciences,vol.,no.,pp.
–, .
[] A. Ducret, M.-P. Valignat, F. Mouhamar, T. Mignot, and
O. eodoly, “Wet-surface-enhanced ellipsometric contrast
microscopy identies slime as a major adhesion factor during
bacterial surface motility,” Proceedings of the National Academy
of Sciences of the United States of America,vol.,no.,pp.
–, .
[]B.Nan,J.Chen,J.C.Neu,R.M.Berry,G.Oster,andD.R.
Zusman, “Myxobacteria gliding motility requires cytoskeleton
rotation powered by proton motive force,” Proceedings of the
National Academy of Sciences of the United States of America,
vol. , no. , pp. –, .
[] C. Wolgemuth, E. Hoiczyk, D. Kaiser, and G. Oster, “How
myxobacteria glide,” Current Biology,vol.,no.,pp.–,
.
[] R. P. Burchard, “Trail following by gliding bacteria,” Journal of
Bacteriology,vol.,no.,pp.–,.
[] R. Yu and D. Kaiser, “Gliding motility and polarized slime
secretion,” Molecular Microbiology,vol.,no.,pp.–,
.
[] E. Ben-Jacob, I. Cohen, and D. L. Gutnick, “Cooperative orga-
nization of bacterial colonies: from genotype to morphotype,”
Annual Review of Microbiology,vol.,pp.–,.
[] P. Vallotton, “Size matters: lamentous bacteria drive interstitial
vortex formation and colony expansion in Paenibacillus vortex,”
Cytometry Part A, vol. , no. , pp. –, .
[] A. B. Holmes, S. Kalvala, and D. E. Whitworth, “Spatial sim-
ulations of myxobacterial development,” PLoS Computational
Biology, vol. , no. , Article ID e, .
[] Y.Wu,Y.Jiang,D.Kaiser,andM.Alber,“Socialinteractionsin
myxobacterial swarming,” PLoS Computational Biology,vol.,
no. , article e, .
[] Y. Wu, A. D. Kaiser, Y. Jiang, and M. S. Alber, “Periodic reversal
of direction allows Myxobacteria to swarm,” Proceedings of the
National Academy of Sciences of the United States of America,
vol. , no. , pp. –, .
[] Y. Wu, N. Chen, M. Rissler, Y. Jiang, D. Kaiser, and M. Alber,
“CA models of myxobacteria swarming,” in Cellular Automata,
vol. , pp. –, Springer, Berlin, Germany, .

Scientica
[] S. S. Branda, ˚
A. Vik, L. Friedman, and R. Kolter, “Biolms: the
matrix revisited,” Tr ends in Micr obi o logy,vol.,no.,pp.–
, .
[] H.-C. Flemming and J. Wingender, “e biolm matrix,” Nature
Reviews Microbiology,vol.,no.,pp.–,.
[] C. B. Whitchurch, T. Tolker-Nielsen, P. C. Ragas, and J. S.
Mattick, “Extracellular DNA required for bacterial biolm
formation,” Science,vol.,no.,article,.
[] J. C. Conrad, “Physics of bacterial near-surface motility using
agella and type IV pili: implications for biolm formation,”
Research in Microbiology,vol.,no.-,pp.–,.
[] J.C.Conrad,M.L.Gibiansky,F.Jinetal.,“Flagellaandpili-
mediated near-surface single-cell motility mechanisms in P.
aeruginosa,” Biophysical Journal,vol.,no.,pp.–,
.
[] M. L. Gibiansky, J. C. Conrad, F. Jin et al., “Bacteria use type IV
pili to walk upright and detach from surfaces,” Science,vol.,
no. , p. , .
[] F. Jin, J. C. Conrad, M. L. Gibiansky, and G. C. L. Wong, “Bacte-
ria use type-IV pili to slingshot on surfaces,” Proceedings of the
National Academy of Sciences of the United States of America,
vol.,no.,pp.–,.
[] K. Zhao, B. S. Tseng, B. Beckerman et al., “Psl trails guide explo-
ration and microcolony formation in Pseudomonas aeruginosa
biolms,” Nature, vol. , no. , pp. –, .
[] S. Wang, M. R. Parsek, D. J. Wozniak, and L. Z. Ma, “A spider
web strategy of type IV pili-mediated migration to build a
bre-like Psl polysaccharide matrix in Pseudomonas aeruginosa
biolms,” Environmental Microbiology,vol.,no.,pp.–
, .
[] M. Schuster, D. Joseph Sexton, S. P. Diggle, and E. Peter
Greenberg, “Acyl-homoserine lactone quorum sensing: from
evolution to application,” Annual Review of Microbiology,vol.
, pp. –, .
[] Y.-H. Li and X. Tian, “Quorum sensing and bacterial social
interactions in biolms,” Sensors,vol.,no.,pp.–,
.
[] A. Ross-Gillespie and R. K¨
ummerli, “Collective decision-mak-
ing in microbes,” Frontiers in Microbiology,vol.,article,
.
[] M. J. Doyle and L. Marsh, “Stigmergy .: from ants to econo-
mies,” Cognitive Systems Research,vol.,pp.–,.
[] E. Bonabeau, “Editor’s introduction: stigmergy,” Articial Life,
vol. , no. , pp. –, .