Sci Nat (2016) 103: 66

DOI 10.1007/s00114-016-1386-8

ORIGINAL PAPER

Migration control: a distance compensation strategy in ants

Thomas A. O’Shea-Wheller 1 & Ana B. Sendova-Franks 2 & Nigel R. Franks 1

Received: 31 May 2016 / Revised: 10 June 2016 / Accepted: 14 June 2016 / Published online: 18 July 2016
# The Author(s) 2016. This article is published with open access at Springerlink.com

Abstract Migratory behaviour forms an intrinsic part of the its robustness lies in the use of simple rules. Additionally, our
life histories of many organisms but is often a high-risk pro- results suggest that such coordinated group reactions are cen-
cess. Consequently, varied strategies have evolved to negate tral to achieving the high levels of ecological success seen in
such risks, but empirical data relating to their functioning are many eusocial organisms.
limited. In this study, we use the model system of the house-
hunting ant Temnothorax albipennis to demonstrate a key Keywords Cost-benefit trade-offs . Decentralised systems .
strategy that can shorten migration exposure times in a group Ecological robustness . Group migration . Temnothorax
of social insects. Colonies of these ants frequently migrate to albipennis
new nest sites, and due to the nature of their habitat, the dis-
tances over which they do so are variable, leading to fluctuat-
ing potential costs dependent on migration parameters. We Introduction
show that colonies of this species facultatively alter the dy-
namics of a migration and so compensate for the distance over Cost-benefit trade-offs are common to almost all life history
which a given migration occurs. Specifically, they achieve this strategies. Animals need to be able to respond to hostile envi-
by modulating the rate of ‘tandem running’, in which workers ronmental conditions such as extremes of temperature and
teach each other the route to a new nest site. Using this meth- predation risk (Creel & Winnie 2005; Hunter 2005; Nonacs
od, colonies are able to engage a larger number of individuals & Dill 1990). Indeed, many life history events involve in-
in the migration process when the distance to be traversed is creased exposure to hazards, but are nevertheless crucial for
greater, and furthermore, the system appears to be based on survival, making a degree of risk essentially unavoidable
perceived encounter rate at the individual level. This form of (Thirgood et al. 2004; Cerdá et al. 1998; Franks & Fletcher
decentralised control highlights the adaptive nature of a be- 1983; Guillemette et al. 2012). As a consequence, strong se-
haviour of ecological importance, and indicates that the key to lection pressures are likely to exist for strategies that mitigate
the dangers of high-risk but necessary behaviours, and a myr-
Communicated by: Sven Thatje iad of such responses may be seen in nature. These range from
Electronic supplementary material The online version of this article
vigilance behaviour in meerkats (le Roux et al. 2009) and
(doi:10.1007/s00114-016-1386-8) contains supplementary material, camouflage in numerous species of mammal (Stevens &
which is available to authorized users. Merilaita 2009) to escape reflexes and group defence in mi-
grating invertebrates (Wine & Krasne 1972; Scho 2005).
* Thomas A. O’Shea-Wheller Social insects also face such challenges, but their responses
to13870@bristol.ac.uk often become manifest at the group or colony level, rather than
at the individual level (Mlot et al. 2011). This enhanced coop-
1
School of Biological Sciences, University of Bristol, Life Sciences erative ability is thought to be a central tenet of their remark-
Building, 24 Tyndall Avenue, Bristol, UK able ecological success (Hunt et al. 2010). A key aspect of the
2
Department of Engineering Design and Mathematics, UWE Bristol, ability of social insect colonies to coordinate group behaviours
Frenchay Campus, Coldharbour Lane, Bristol, UK is the need for reliable and rapid information relating to the

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

66 Page 2 of 9 Sci Nat (2016) 103: 66

immediate environment, with examples encompassing sens- distances are likely to increase the time that workers spend travel-
ing of temperature (Rosengren et al. 1987), perception of the ling between nests, prolonging their exposure period, and thus the
intensity of predation (O’Shea-Wheller et al. 2015; Whitford degree of risk (Langridge et al. 2008). Hence, we predict that
& Bryant 1979) and the detection of potentially hostile con- colonies will be able to compensate for distance. To assess any
specifics (Franks & Fletcher 1983). Another factor that influ- such strategies, in our experiments, we provide colonies with both
ences colony exposure to risk is the distance that must be ‘near’ and ‘far’ new nest sites and allow them to migrate. During
travelled either as part of foraging, or during migration behav- migrations, their current nests are left intact, acting to prolong the
iours. Distance-cost scaling can be observed in vertebrate for- decision-making process and allowing us to measure more finely
aging (Lima 1985), and is likely also to apply to social insects, the dynamics of colony relocation (Dornhaus et al. 2004).
as the time that individuals spend away from the safety of their In this study, we focus on several key migration behaviours.
nest necessarily increases as a function of distance travelled First, we record the amount of forward and reverse tandem
(Traniello 1989). Consequently, many species of social insects running conducted by colonies. Tandem running involves in-
have evolved methods of estimating distances to food sources formed workers leading naïve colony mates to a new nest site
and potential new nest sites. For example, there is evidence (forward tandem running), or from the new site back to their old
that bees measure distance using both optic flow odometry nest (reverse tandem running), in order to teach them the rela-
and path integration, helping them to find desirable resources tive locations of each (Franks & Richardson 2006). Tandem
and relay their location to other colony members (Srinivasan runs are crucial to the progression of migrations, as movement
et al. 1997; Chittka et al. 1995). Within the Formicidae, the requires an active ‘corps’ of spatially informed workers (Franks
desert-dwelling ant Cataglyphis fortis also uses path integra- et al. 2009). Second, we quantify aspects of quorum sensing.
tion, but based on a combination of celestial compass cues and This is a process in which ants assess the number of nest mates
pedometry, allowing reliable navigation in its expansive and within a potential new nest site, and once the population in a
relatively featureless environment (Wittlinger et al. 2006; candidate site reaches a certain value, termed the ‘quorum
Wehner 2009). However, there are fewer studies relating to threshold’ ants will switch from slow tandem running to rapid
how social insects may use such information to reduce mor- social carrying, a transition known as ‘quorum attainment’
tality and risk of disorientation at the group level. (Franks et al. 2015). Quorum sensing, by which ants assess
The ant Temnothorax albipennis is a model species that is the quorum threshold, is thus central to solidifying a colony’s
well suited to testing how an increase in perceived risk may target nest choice and in implementing the subsequent migra-
influence colony-level behaviour. Colonies of this species live tion process. Third, we measure the per capita rate of social
within fragile rock cavities in a highly heterogeneous environ- carrying, in which workers, informed of a new nests location
ment, and will migrate into better structures under both field via tandem running, rapidly carry their colony mates to the new
and laboratory conditions, whenever the opportunity is present- nest site, once a quorum threshold has been achieved (Planqué
ed. Colony migrations to new nest sites thus make up a crucial et al. 2007). Social carrying allows migrations to proceed rap-
part of the species’ life history, but at the same time represent a idly after a decision has been made, as it is three times faster
high-risk process, whereby the vulnerable brood and all colony than tandem running (Planqué et al. 2007).
members are exposed (Dornhaus et al. 2004). Furthermore, the We hypothesise that when nest sites are further away, col-
distance between the old and new nests influences total migra- onies will increase the use of behaviours implicated in the
tion rate and thus is inseparably linked to the total level of risk rapidity of migrations, such as forward and reverse tandem
during any given migration. Migration distances are highly running. Furthermore, it may be postulated that over longer
variable; colonies will migrate into a better quality nest even distances, workers might employ a lower quorum threshold, in
if it is as far as 2.85 m from their current one, or as close as order to commit to a new nest more rapidly, begin carrying
10 mm (Franks et al. 2008). As such, it seems likely that T. nest mates sooner, and thus expedite the migration. By analy-
albipennis colonies employ strategies to minimise risk when sis of such metrics, in concert with migration rates over dif-
conducting migrations over longer distances, and there are sev- ferent distances, we aim to assess how, and indeed whether,
eral candidate behaviours that may be involved in this. Perhaps colonies make adjustments in order to counteract the increased
most important in ensuring the continuity and speed of migra- danger posed by longer migrations.
tions is a behaviour known as tandem running. This involves an
ant that knows the location of the new nest leading another
colony member to that nest and thus teaching it the route, and Materials and methods
is implicated in migration rate and distance assessment
(Hölldobler & Wilson 1990; Franks & Richardson 2006). Colonies
Here, using migrating colonies of T. albipennis, we aim to test
whether the distance that must be travelled to a new nest affects the We collected 30 colonies of T. albipennis from the Isle of
expression of important migration behaviours. Longer migration Portland, Dorset (50.547889°N, −2.448251°W), on 22

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Sci Nat (2016) 103: 66 Page 3 of 9 66

January 2015. Each contained a queen, from 45 to 397 Experiment protocol

workers, and 11 to 356 brood items (eggs, larvae and pupae).
No permission for collection was required as T. albipennis is A total of 30 emigrations were conducted over a 19-day period
not a protected species and colonies were taken from a dis- from the 2nd to the 20th of February, 2015. Experiments were
turbed quarry area with free public access. However, care was performed in 1900 × 1000-mm plastic arenas with Fluon-
taken to keep disturbance of the population to a minimum by coated sides (Fig. 1). These large arenas were built in order
use of a rota based on the dates and locations of previous to ensure an extensive potential scouting area, and thus better
collections. Colonies were maintained under standard labora- simulate natural conditions, as ants of this species have been
tory conditions, and all nests were housed in plastic Petri known to migrate into new nest sites up to 2850 mm away
dishes with Fluon-coated sides to avoid worker escape. Ants (Franks et al. 2008). Two trials were conducted each day si-
were fed with Drosophila melanogaster weekly, and allowed multaneously, in order to limit the total experimental duration.
to forage for water and honey solution ad libitum in accor-
dance with established upkeep methods (Dornhaus et al.
2004).
Nests of two different qualities (Franks et al. 2003)
were used in the experiments, both with an internal cavity
area of 60 × 35 mm. ‘Poor’ quality nests, in which all
colonies were initially housed, had 1-mm-high walls, a
6-mm-wide entrance and a clear cover. ‘Excellent’ quality
nests, into which colonies were able to migrate during
trials, had 2-mm-high walls, a 1-mm-wide entrance and a
red filter cover. The poor nests had transparent lids to
simulate cavities that were compromised by high light
levels. The use of red filters in excellent nests ensured
lower perceived light levels (Ogawa et al. 2015). This
made them preferable to migrating ants, but at the same
time did not obscure the observation of individuals within
these nests (Franks et al. 2006).

Treatment groups

After collection, all colonies were allowed to migrate into

initial poor quality nests and left to acclimatise for a period
of 7 days. This acclimatisation interval was used as colonies
gain experience with multiple migrations, leading to changes
in decision speed, and such experience is lost only after 7 days
of inactivity (Langridge et al. 2004). Prior to initiation of the
experiments, we grouped each of the 30 colonies into sets of 3
based on size, then randomly assigned each colony to one of
the following three experimental groups: the ‘near nest’
group, the ‘distant nest’ group and the ‘delayed near nest’
group. Each colony underwent only the experimental trial to
which it was assigned and emigrated only once. New nests in
the ‘delayed near’ treatment were no further away than in the
‘near’ treatment but were only placed into arenas after a sub-
stantial time interval of 90 min. This 90-min interval was Fig. 1 Layout of experimental arenas for a the near treatment; b the near
chosen as it allowed time for scouts to reach and explore the delayed treatment, in which the excellent new nest was only placed into
area in which the new nest would eventually be placed. The the arena after 90 min; and c the distant treatment. Starting nests housing
purpose of this treatment was to assess whether the time a Temnothorax albipennis ant colonies were of poor quality and were not
destroyed, allowing voluntary emigrations to occur. The search angles
colony spent scouting in an area before a new nest site was leading to discovery of new nests are indicated; all new nest sites were
found affected the perceived accessibility of such a nest, and of excellent quality. Figure aims to demonstrate nest quality setup and
thus elicited changes in migration behaviour. thus is not to scale

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

66 Page 4 of 9 Sci Nat (2016) 103: 66

This was desirable, as T. albipennis colonies show seasonality, were included in the analyses, but no data were recorded
and thus, shorter experimental periods are required for behav- after this point.
ioural consistency (Franks et al. 2006). The order of treat-
ments used was also randomised to make sure that no unin- Data analysis
tended time-related biases occurred. Prior to each trial, the
arenas were cleaned with alcohol and water to remove any In the analysis of all factors, Shapiro-Wilk tests were
odours, and lit evenly using an overhead LED lighting unit employed to determine if the data differed significantly from
(colour rendering index 85, luminous flux 4100). a normal distribution. For normally distributed data with ho-
Occupied poor nests were placed into fixed positions in mogeneous variance, we used one-way ANOVA with Tukey’s
the arenas, and new excellent nests were placed at set dis- HSD post hoc tests to analyse differences between treatment
tances from these, dependent on the treatment group. For groups. Where the data was found to be not normally distrib-
the near nest group, new nests were positioned so that their uted, we transformed it using either arcsine square root, or
entrance was 100 mm from the starting nest, whilst for the log10 transformations, dependent on its skewness. This was
distant nest group, new nests were set up at a distance of done to ensure homogeneity of statistical testing methods.
300 mm (Fig. 1a, c). The potential search angles that Means quoted from this data were reverse transformed after
allowed discovery of these nests were measured by extrap- calculation. Transformations did not alter the pattern of statis-
olating the front face size of new nests over the distance to tical significance, confirmed by testing data before and after
the original nest entrance and then quantifying the angle transformation, with both parametric and non-parametric tests.
with a protractor (Fig. 1). New nests were placed into the For per capita forward and reverse tandem running, relative
open arenas as the use of bridges or channels would artifi- quorum threshold and the per capita rate of social carrying,
cially lead ants towards them, and thus remove any poten- data were normalised for colony size. This was achieved by
tial differences in nest discovery frequency. The delayed calculating values in proportion to colony population, in order
near nest treatment was also set up with a 100-mm separa- to determine the per capita expression of a behaviour based on
tion to ensure consistency. In the delayed near treatment, the number of available individuals.
the new nest was only placed into the arena 90 min after We used a Kaplan-Meier survival analysis to assess the
the nest occupied by the colony, and recording began only effect of treatment group on the mean percentage of ants and
after the 90-min lag period (Fig. 1b). During the movement brood building up in new nests over time. We used means of
of nests into the arenas, any scouts that were outside their the data to produce a cumulative survival function, and ants
nests in the original Petri dishes were placed onto the tops and brood that moved after 180 mins were treated as censored
of the starting nests within the arenas, using a fine paint points. Pairwise comparisons were made using the Mantel-
brush to minimise disturbance. This was done in order to Cox test, for which alpha was 0.016 based on a Bonferroni’s
avoid any scouting discrepancies caused by different initial correction for the three possible pairwise comparisons be-
locations of scouts. From this point, we allowed emigra- tween treatment groups. All analyses were conducted in
tions to proceed whilst recording the number of workers SPSS (Release version 21.0.0.0, IBM Corporation and
and brood in each nest every 10 min for the first other(s) 1989, 2012).
180 min. We also recorded, via direct counts, the timings
of new nest discovery, first tandem run, quorum attainment
(measured by first instance of social carrying) and the Results
amount of time that first discoverers spent in each new
nest. Additionally, we quantified the per capita number of Behavioural phases and tandem running
forward and reverse tandem runs, relative quorum threshold
and per capita number of worker and brood social carries, The timings of nest discovery (one-way ANOVA,
with tally counts. Relative rather than absolute quorum F 2,27 = 17.439, P < 0.001), first tandem run (one-way
threshold was used for consistency and comparability, as ANOVA, F2,27 = 13.384, P < 0.001) and quorum attainment
quorum threshold scales linearly with colony size (one-way ANOVA, F2,27 = 7.543, P = 0.023) were significant-
(Dornhaus and Franks 2006). As migrations were generally ly affected by treatment group. This was due to the timings of
slow and all behaviours were distinct, recordings were n e s t d i s c o v e r y ( m e a n s n e a r = 11 . 9 4 m i n , n e a r
made manually, in accordance with established laboratory delayed = 6.66 min, distant = 14.15 min), first tandem run
techniques (O’Shea-Wheller et al. 2015). If migrations were (means near = 32.80 min, near delayed = 19.74 min,
not completed within 180 min, colonies were left for a distant = 57.82 min) and quorum attainment (means
further 24 h before returning them to their holding dishes. near = 69.26 min, near delayed = 54.12 min, distant =
For incomplete migrations, all relevant behaviours that oc- 115.57 min) within the near nest and delayed near nest
curred up to the end of the 180-min observation period treatments occurring significantly earlier than in the

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Sci Nat (2016) 103: 66 Page 5 of 9 66

distant nest treatment. Principally, these results can be 1.184 %) and the number of reverse tandem runs per capita
explained by the increased difficulty of finding new (one-way ANOVA, F2,27 = 0.355, P = 0.704) also did not dif-
nests in the distant treatment and consequent slowing fer significantly between treatments.
of migration progression.
The per capita number of tandem runs also varied signifi- Progression of migrations
cantly between groups (one-way ANOVA, F2,27 = 4.559,
P = 0.020). Specifically, there was a higher per capita number The average percentage of ants building up over time in the
of tandem runs in the distant nest treatment than in either the new nests was significantly different between groups (Mantel-
near nest (Tukey’s HSD, mean difference = −0.106, P = Cox test statistic = 17.349, df = 2, P < 0.001). Specifically, this
0.048) or delayed near nest (Tukey’s HSD, mean difference = was explained by differences between the near nest and distant
−0.115, P = 0.030) treatments (Fig. 2). This was likely caused nest treatments (Mantel-Cox test statistic = 13.708, df = 1,
by a lower rate of independent nest discovery by workers in P < 0.001) and delayed near nest and distant nest treatments
the distant treatment; thus, more tandem runs were needed to (Mantel-Cox test statistic = 15.700, df = 1, P < 0.001).
reach a quorum threshold in this group. However, the near nest and delayed near nest treatments were
The relative quorum threshold (one-way ANOVA, not significantly different (Mantel-Cox test statistic = 0.375,
F2,27 = 1.240, P = 0.305), per capita rate of social carrying df = 1, P = 0.540; Fig. 3). This suggests that the longer journey
(one-way ANOVA, F2,27 = 0.570, P = 0.572) and time that to nests in the distant treatment lead to a reduction in migration
first discoverer spent in nest (one-way ANOVA, F2,27 = rate, but that this was not the case in either the near or delayed
1.179, P = 0.323) did not differ significantly between treat- near treatments, both being equidistant from the original nest.
ments. Such a consistent relative quorum threshold is contrary The average percentage of brood building up in the new
to our hypothesis that lower thresholds may be employed over nest over time was also significantly different between groups
longer distances, and may be accounted for by the move-to- (Mantel-Cox test statistic = 69.132, df = 2, P < 0.001).
improve conditions used in our experiments (Dornhaus et al. Significant differences occurred between the near nest and
2004). Furthermore, the lack of change in social carrying rates distant nest treatments (Mantel-Cox test statistic = 46.416,
may possibly be explained by the comparable relative quorum df = 1, P < 0.001) and delayed near nest and distant nest treat-
thresholds between treatments. Similar quorums suggest that ments (Mantel-Cox test statistic = 72.298, df = 1, P < 0.001).
the average numbers of workers in the new nests at the begin- However, again the near nest and delayed near nest treatments
ning of carrying were similar in each treatment, and these were not significantly different (Mantel-Cox test statistic =
workers were then the ones that contributed to the carrying 3.375, df = 1, P = 0.066; Fig. 3). These results point to the
process. As such, numbers of carriers were unlikely to differ same trend as seen in worker movement, whereby only in-
between treatments, despite changes in tandem running. creased distance led to a slower migration. Notably, overall
Lastly, as new nest qualities were always excellent, it is not brood transport was faster than the movement of workers.
necessarily surprising that discoverers spent a consistent time This is because brood items are immobile, and due to their
within them, regardless of distance. The percentage of ants in vulnerability, are rapidly moved to the new nest with prefer-
the new nest at the time of first tandem run (initial independent ence over adults (Langridge et al. 2008).
discoveries; one-way ANOVA, F2,27 = 1.441, P = 0.254; The percentages of workers (one-way ANOVA,
means near = 2.750 %, near delayed = 2.165 %, distant = F2,27 = 2.798, P = 0.079) and brood (one-way ANOVA,

Fig. 2 Mean number of tandem

runs (per capita) across
treatments. Significant
differences between groups
(P < 0.05), based on ANOVA
analysis and Tukey’s HSD post
hoc tests, are indicated with
asterisks. N = 10 colonies were
used in each treatment; error bars
indicate 95 % confidence
intervals

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

66 Page 6 of 9 Sci Nat (2016) 103: 66

Fig. 3 Mean percentage of total

colony workers (a) and brood (b)
in the new nests over time for
each of the three treatment
groups. Blue lines indicate the
near treatment, grey lines indicate
the near delayed treatment and
black lines indicate the distant
treatment. In the near delayed
treatment, colonies were allowed
to scout for 90 min but with no
new nest present until time 0. The
mean represents n = 10 colonies
for each time point; asterisks
indicate the mean time of first
brood transport (quorum
attainment), in each
correspondingly coloured group.
All error bars indicate 95 %
confidence intervals

F2,27 = 3.178, P = 0.058) in the new nests at the final time nest is further away, in order to expedite the migration process,
points did not differ significantly between treatment groups, by increasing the rate at which workers discover the new nest
but it must be noted that values in the distant treatment were (Stuttard et al. 2015). Our results demonstrate that travel dis-
still lower for both measures, and both p values were small tance significantly affects colony migration strategy in move-
even if above 0.05 (Fig. 4). This is indicative of a partial, but to-improve scenarios, and additionally, they highlight the im-
incomplete, adjustment of migration rates by the end of the portance of perception at the individual level in influencing
process. collective behaviour.
Independent discoveries—that is ants finding the new nest
without being led to it—were less frequent in the far nest
Discussion treatment than in the near or near delayed treatments. This
was characterised by the buildup of ants in the new nests prior
We show that colonies of T .albipennis can respond to differ- to the initiation of tandem running, as all individuals present at
ing migration conditions via the regulation of tandem running this time made independent discoveries. The numbers of ants
behaviour. Specifically, we found that colonies will increase in new nests at the time of the first tandem run did not differ
the per capita number of tandem runs used during an emigra- significantly between treatments, suggesting that tandem run-
tion when the distance to be traversed is greater (Fig. 2). In this ning began at a set nest population. However, crucially, along
way, migration transit time is minimised and exposure period with quorum attainment, the first tandem run occurred later in
is reduced. Furthermore, we found that the end result of mi- the distant nest treatment (Fig. 3). This showed that more time
grations after 180 min across colonies was comparable regard- was required to reach such a population and was indicative of
less of the distance travelled (Fig. 4), but that the underlying a lower discovery rate when the nest was further away.
dynamics varied significantly (Figs. 2 and 3). This indicates Furthermore, the search angle leading to discovery of the front
that colonies increase the rate of tandem running when a new surfaces of the new nests in the near and delayed near

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Sci Nat (2016) 103: 66 Page 7 of 9 66

Fig. 4 Boxplots for percentages of ants (a) and brood (b) in the new nests lower quartiles, respectively. Outliers (further than 1.5 times the
at 180 min, n = 10 colonies per treatment. Boxes are cut at the median, interquartile range from the median) are marked with circles (ants) and
with the upper and lower limits of the box representing the upper and asterisks (brood); numbers indicate colony IDs

treatments (40°), was more than double that of the distant explanation for our results (Pratt 2004). When nests are near-
treatment (15°; Fig. 1), indicating that distant nests were by, exploring workers are likely to find them relatively easily.
harder to find, and additionally presented smaller vertical However, this is not the case when they are distant. Hence,
landmarks. Indeed, such differences are ecologically relevant more tandem runs are required over longer migration dis-
to the probability of nest discovery, principally because ants of tances simply because independent discoveries are rarer, and
this species do use visual landmarks, and upon contact with thus, more ants must be led to a nest before a quorum thresh-
the front of a nest, scouting workers will often employ thig- old can be reached. In this way, workers do not measure dis-
motaxis to locate its entrance (Basari et al. 2014; Hunt et al. tance directly, but instead respond to quorum attainment, a
2014). factor that in this case was influenced by distance (Pratt
As distant nests were more difficult to discover, the role of 2004). It is known that T. albipennis workers use encounter
increased numbers of tandem runs was clear; they led to an frequency to measure nest size and quorum threshold (Stuttard
increase in the discovery rate and thus generated a larger pool et al. 2015; Pratt 2004) and that tandem running will continue
of informed workers, potentially able to contribute to the car- up until the point that a quorum threshold is reached (Planqué
rying stage of the migration (Franks & Richardson 2006; et al. 2007). Lower discovery rates extend the pre-quorum
Planqué et al. 2007). Conversely, the absence of this process period, as seen in the distant treatment, and so a greater pro-
in the delayed near nest treatment can be explained because portion of ants in the new nest are led there by tandem runs. As
such nests were no further away than in the near treatment. such, the greater number of tandem runs is likely a self-
Without an increased travel distance, the independent discov- regulating process, controlled by quorum attainment alone,
ery rate and time taken to recruit naive ants to the new nest rather than by actual distance measurement.
would not have differed between workers in the two near The aforementioned self-organisation hypothesis is consis-
treatments. Hence, there was little stimulus for colonies to tent with earlier work on T.albipennis, T.curvispinosus and
compensate based on perception at the individual level, de- T.rugatulus, suggesting that when independent nest discover-
spite delayed nest discovery. This is advantageous, as in their ies were more common, the rate of tandem running decreased
natural habitat, nest discovery times are likely to be more (Mallon et al. 2001; Pratt 2008). Notably, whilst these studies
variable due to non-uniform environmental features. In light used forced migrations, involving the destruction of colonies’
of this, it is reasonable to assume that colonies accounted for original nests, our own data appear to corroborate their find-
the increased distance between nests without the use of dis- ings in relation to environmentally common move-to-improve
covery time as a proxy. migrations, in which tandem running plays a greater role
Although the use of landmarks and pedometry have been (Dornhaus et al. 2004). Tandem runs to near nests are rare or
implicated in nest location during tandem running (Franklin & absent under emergency migration conditions (Dornhaus et al.
Franks 2012; Basari et al. 2014), here, a system based on 2004), but in our experiments, all colonies in the near treat-
independent discovery rate provides a more parsimonious ment still engaged in this behaviour. Indeed, there is a clear

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

66 Page 8 of 9 Sci Nat (2016) 103: 66

advantage to doing so under benign conditions, as during the systems (Siljak 2012; Zimmerman et al. 2014), the key to this
tandem running stage, colonies are still open to switching to a behaviour lies in its decentralised control. Thus, we provide
better nest site if one is discovered, whilst this is no longer the further evidence that the robustness of self-organised systems
case once they commit to carrying (Franks et al. 2007; Sasaki is one of the central factors enabling eusocial animals to meet
et al. 2015). ecological challenges.
There is a substantial body of literature implying that for
any given migration, quicker is better (Langridge et al. 2008; Author contributions TAOW and NRF conceived the study and wrote
the manuscript, TAOW conducted all the experimental work, and TAOW
Scharf et al. 2012; Franks et al. 2009; Franks et al. 2008), and
and ABSF carried out the statistical analysis. All authors gave final ap-
colonies in our experiment did attempt to expedite migrations. proval for publication.
In spite of this, we found that longer migration distances still .
led to slower movement, even with the partial rescuing of Compliance with ethical standards
migration rate by increased tandem running (Figs. 3 and 4).
This highlights the often-incomplete nature of compensatory Conflicts of interest The authors declare that they have no conflict of
interest.
behaviours, and suggests that for T.albipennis colonies, in-
creased migration distance may always impart some degree Data accessibility All data are included in the Electronic supplemen-
of cost. tary material.
Whilst tandem running is a crucial behaviour implicated in
Ethical standard statement All applicable institutional and/or national
regulating migrations (Franks & Richardson 2006), there is guidelines for the care and use of animals were followed.
also evidence to suggest that reverse tandem runs; from the
new to the old nest, play a role in modulating migration rate Funding This work received no specific funding.
(Franks et al. 2009). However, we found no significant effect
of migration distance on the number of reverse tandem runs. Open Access This article is distributed under the terms of the Creative
This result may be explained by the findings of previous em- Commons Attribution 4.0 International License (http://
pirical studies, showing that increased numbers of reverse creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
tandem runs are induced only by disorientation of active distribution, and reproduction in any medium, provided you give
appropriate credit to the original author(s) and the source, provide a link
scouts. Such studies indicate that the primary purpose of re- to the Creative Commons license, and indicate if changes were made.
verse tandem running is to re-engage ‘lost scouts’ in the mi-
gration process, rather than to train naïve workers (Planqué
et al. 2007; Franks et al. 2009). In our experiments, distance
References
was the key limiting factor; thus, it was preferable to engage as
many new workers in the tasks of recruitment and carrying, a
Basari N, Bruendl AC, Hemingway CE, Roberts NW, Sendova-Franks
process achieved by the use of forward tandem runs alone AB, Franks NR (2014) Landmarks and ant search strategies after
(Franks & Richardson 2006). As such, it is probable that col- interrupted tandem runs. J Exp Biol 217(Pt 6):944–954
onies facultatively alter both forward and reverse tandem run- Cerdá X, Retana J, Manzaneda A (1998) The role of competition by
ning to control migration rate, dependent on the particular dominants and temperature in the foraging of subordinate species
in Mediterranean ant communities. Oecologia 117(3):404–412
challenges that threaten to curtail it. Chittka L, Kunze J, Shipman C, Buchmann SL (1995) The significance
Building on the work of Pratt (2004), we show the potential of landmarks for path integration in homing honeybee foragers.
for colonies to modulate their behaviour in response to dis- Naturwissenschaften 82(7):341–343
tance, using a decentralised system based on perception at the Creel S, Winnie J a (2005) Responses of elk herd size to fine-scale spatial
individual level. Indeed, it may well be argued that for the and temporal variation in the risk of predation by wolves. Anim
Behav 69(5):1181–1189
purposes of migration in T.albipennis, such a comparatively Dornhaus A, Franks NR (2006) Colony size affects collective decision-
simple system provides superiority in both rapidity and ro- making in the ant Temnothorax albipennis. Insectes Soc 53(4):420–
bustness when juxtaposed against more sophisticated distance 427
assessment methods (Thiélin-Bescond & Beugnon 2005). Dornhaus A, Franks NR, Hawkins RM, Shere HNS (2004) Ants move to
improve: colonies of Leptothorax albipennis emigrate whenever
Furthermore, the specific costs addressed by this process are
they find a superior nest site. Anim Behav 67(5):959–963
of particular ecological significance, as migration distances Franklin EL, Franks NR (2012) Individual and social learning in tandem-
vary widely in nature (Franks et al. 2008). running recruitment by ants. Anim Behav 84(2):361–368
Our work demonstrates an example of colony-level risk Franks NR, Fletcher CR (1983) Spatial patterns in army ant foraging and
mitigation behaviour, which, taken with other such examples migration : Eciton burchellii on Barro Colorado Island, Panama.
Behav Ecol Sociobiol 12(4):261–270
(O’Shea-Wheller et al. 2015; Gordon et al. 2007), further il-
Franks NR, Richardson T (2006) Teaching in tandem-running ants.
lustrates the enhanced ability of social insects to adapt and Nature 439(7073):153
respond to risks over a range of spatial and temporal scales. Franks NR, Mallon EB, Bray HE, Hamilton MJ, Mischler TC (2003)
Moreover, as is the case with numerous other complex Strategies for choosing between alternatives with different

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Sci Nat (2016) 103: 66 Page 9 of 9 66

attributes: exemplified by house-hunting ants. Anim Behav 65(1): O’Shea-Wheller TA, Sendova-Franks AB, Franks NR (2015)
215–223 Differentiated anti-predation responses in a superorganism. PLoS
Franks NR, Dornhaus A, Metherell BG, Nelson TR, Lanfear SAJ, Symes One 10(11):e0141012
WS (2006) Not everything that counts can be counted: ants use Ogawa Y, Falkowski M, Narendra A, Zeil J, Hemmi JM (2015) Three
multiple metrics for a single nest trait. Proc R Soc Lond Biol spectrally distinct photoreceptors in diurnal and nocturnal Australian
273(1583):165–169 ants. Proc R Soc Lond Biol 282(1808)
Franks NR, Hooper JW, Gumn M, Bridger TH, Marshall JAR, Groß R, Planqué R, Dechaume-Moncharmont F-X, Franks NR, Kovacs T,
Dornhaus A (2007) Moving targets: collective decisions and flexible Marshall JAR (2007) Why do house-hunting ants recruit in both
choices in house-hunting ants. Swarm Intell 1:81–94 directions? Naturwissenschaften 94(11):911–918
Franks NR, Hardcastle KA, Collins S, Smith FD, Sullivan KME, Pratt SC (2004) Quorum sensing by encounter rates in the ant
Robinson EJH, Sendova-Franks AB (2008) Can ant colonies choose Temnothorax albipennis. Behav Ecol 16(2):488–496
a far-and-away better nest over an in-the-way poor one? Anim Pratt SC (2008) Efficiency and regulation of recruitment during colony
Behav 76(2):323–334 emigration by the ant Temnothorax curvispinosus. Behav Ecol
Franks NR, Dechaume-Moncharmont FX, Hanmore E, Reynolds JK Sociobiol 62(8):1369–1376
(2009) Speed versus accuracy in decision-making ants: expediting
Rosengren R, Fortelius W, Lindström K, Luther A (1987) Phenology and
politics and policy implementation. Phil Trans R Soc B 364(1518):
causation of nest heating and thermoregulation in red wood ants of
845–852
the Formica rufa group studied in coniferous forest habitats in
Franks NR, Stuttard JP, Doran C, Esposito JC, Master MC, Sendova-
southern Finland. Ann Zool Fenn 24(2):147–155
Franks AB, Masuda N, Britton NF (2015) How ants use quorum
sensing to estimate the average quality of a fluctuating resource. Sci Sasaki T, Colling B, Sonnenschein A, Boggess MM, Pratt SC (2015)
Rep 5:11890 Flexibility of collective decision making during house hunting in
Gordon DM, Holmes S, Nacu S (2007) The short-term regulation of Temnothorax ants. Behav Ecol Sociobiol 69(5):707–714
foraging in harvester ants. Behav Ecol 19(1):217–222 Scharf I, Modlmeier AP, Fries S, Tirard C, Foitzik S (2012)
Guillemette M, Richman SE, Portugal SJ, Butler PJ (2012) Behavioural Characterizing the collective personality of ant societies: aggressive
compensation reduces energy expenditure during migration hyper- colonies do not abandon their home. PLoS One 7(3):e33314
phagia in a large bird. Behav Ecol 26(4):876–883 Scho C (2005) Temporal and spatial patterns in the emigrations of the
Hölldobler B, Wilson EO (1990) The Ants. The Belknap press of Harvard army ant Dorylus (Anomma) molestus in the montane forest of Mt.
University Press, Cambridge, Massachusetts Kenya. Ecol Entomol 30:532–540
Hunt BG, Wyder S, Elango N, Werren JH, Zdobnov EM, Yi SV, Siljak DD (2012) Decentralized control of complex systems. Dover
Goodisman MAD (2010) Sociality is linked to rates of protein evo- Publications Inc., Mineola, New York
lution in a highly social insect. Mol Biol Evol 27(3):497–500 Srinivasan MV, Zhang SW, Bidwell NJ (1997) Visually mediated
Hunt ER, O’Shea-Wheller T, Albery GF, Bridger TH, Gumn M, Franks odometry in honeybees. J Exp Biol 200:2513–2522
NR (2014) Ants show a leftward turning bias when exploring un- Stevens M, Merilaita S (2009) Animal camouflage: current issues and
known nest sites. Biol Lett 10(12):20140945 new perspectives. Phil Trans R Soc B 364(1516):423–427
Hunter LM (2005) Migration and environmental hazards. Popul Environ Stuttard JP, Gottlieb D, Franks NR (2015) Ants incommunicado: collec-
26(4):273–302 tive decision-making over new nest sites by ants with reduced com-
Langridge EA, Franks NR, Sendova-Franks AB (2004) Improvement in munication. Behav Ecol Sociobiol 70(1):145–155
collective performance with experience in ants. Behav Ecol Thiélin-Bescond M, Beugnon G (2005) Vision-independent odometry in
Sociobiol 56(6):523–529 the ant Cataglyphis cursor. Naturwissenschaften 92(4):193–197
Langridge EA, Sendova-Franks AB, Franks NR (2008) The behaviour of Thirgood S, Mosser A, Tham S, Hopcraft G, Mwangomo E, Mlengeya T,
ant transporters at the old and new nests during successive colony Kilewo M, Fryxell J, Sinclair ARE, Borner M (2004) Can parks
emigrations. Behav Ecol Sociobiol 62(12):1851–1861 protect migratory ungulates? The case of the Serengeti wildebeest.
Le Roux A, Cherry MI, Gygax L, Manser MB (2009) Vigilance behav- Anim Conserv 7(2):113–120
iour and fitness consequences: comparing a solitary foraging and an
Traniello JFA (1989) Foraging strategies of ants. Annu Rev Entomol
obligate group-foraging mammal. Behav Ecol Sociobiol 63(8):
34(1):191–210
1097–1107
Lima SL (1985) Maximizing feeding efficiency and minimizing time Wehner R (2009) The architecture of the desert ant’s navigational toolkit
exposed to predators: a trade-off in the black-capped chickadee. (Hymenoptera: Formicidae). Myrmecol News 12(September):85–
Oecol 66(1):60–67 96
Mallon EB, Pratt SC, Franks NR (2001) Individual and collective Whitford WG, Bryant M (1979) Behavior of a predator and its prey: the
decision-making during nest site selection by the ant Leptothorax horned lizard (Phrynosoma cornutum) and harvester ants
albipennis. Behav Ecol Sociobiol 50(4):352–359 (Pogonomyrmex spp.). Ecology 60(4):686–694
Mlot NJ, Tovey C a, Hu DL (2011) Fire ants self-assemble into water- Wine BYJJ, Krasne FB (1972) The organization of escape behaviour in
proof rafts to survive floods. Proc Natl Acad Sci U S A 108(19): the crayfish. J Exp Biol 56:1–18
7669–7673 Wittlinger M, Wehner R, Wolf H (2006) The ant odometer: stepping on
Nonacs P, Dill LM (1990) Mortality risk vs. food quality trade-offs in a stilts and stumps. Science 312:1965–1967
common currency: ant patch preferences. Bull Ecol Soc Am 71(5): Zimmerman LM, Bowden RM, Vogel LA (2014) A vertebrate cytokine
1886–1892 primer for eco-immunologists ed. Funct Ecol 28(5):1061–1073

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:

1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at

onlineservice@springernature.com