Behavioural Processes, 31 (1994) 0 1994 Elsevier Science B.V. All BEPROC
111-I rights
24 reserved
111 0376-6357/94/$07.00
00516
Trail-laying foraging
and recruitment
red wood-ants Formica aquilonia Yarrow (
Hymenoptera: A.E.
University
to sugary foods by
of Aberdeen,
Formicidae)
Lamb * and J.G. Ollason
Department of Zoology, Culterty Field Station, Aberdeenshire AB4 I OAA, UK
Newburgh,
El/on,
(Accepted 20 August 1993)
Abstract The use of trail communication by red wood-ants, Formica aquilonia Yarrow, foraging to sugar baits, was investigated by providing a laboratory colony with aqueous sucrose at two novel feeding sites, with a third site, providing no food, as a control. During trials, the sites were connected to the nest by paper bridges which were replaced between trials. One of the sites providing food was reached by a two-way bridge, as was the control. The second baited site was reached by a one-way bridge: this site had a one-way exit from which returning foragers dropped to the nest. Hence, if returning foragers laid a trail from a bait, they would do so on the two-way bridge, but not on the one-way bridge by which foragers did not return, nor on the bridge to the unbaited site. This would lead to a greater increase in foraging traffic on the two-way bridge, relative to that on the other two bridges, during a given trial. Each trial was preceded by a 30-min counting session with no baits present, to establish the mean number of ants arriving per minute on each bridge through random exploration. After provision of food, ants arriving on each bridge per minute were counted in four separate 30-min sessions: each count was then converted to a proportion of the original mean traffic on that bridge, and the change in traffic with time was compared between bridges using analysis of variance. The increase in traffic was significantly greater on the two-way bridge than on the control and one-way bridges, indicating that replete ants returning from the bait left a trail on the homeward route, which other ants subsequently followed.
Key words:
Foraging; Recruitment;
* Corresponding
SSDI
author.
0376.6357(93)E0046-5
Ant; Formica
aquilonia
112
Introduction Among red wood-ants (Formica rufa L. and allied species), communication between foragers takes place against a background of independent decisions by individual ants. Although they are generalist predators, wood-ants are specialists on honeydew from aphids as a carbohydrate source (Carter and Maslen, 1982; Rosengren and Sundstrom, 1991). Clones of aphids are predictably located on trees within a colony’s territory, and individual foragers readily learn the route to a particular feeding site, subsequently becoming constant to it (Rosengren, 1971, 1977; Cosens and Toussaint, 1985, 1986; Rosengren and Fortelius, 1986; Whittaker, 1991; Gordon et al., 1992). Hence, when aphids are plentiful, these ants may be able to rely upon memory to obtain sufficient carbohydrate for their needs, without exchanging directional information. Until the present study, trail recruitment to sugary foods has proved difficult to demonstrate in F. aquilonia, apparently because learning and site fidelity tend to obscure its effects (Cosens and Toussaint, 1985). Nevertheless, directional recruitment has been observed in F. polyctena, whose foragers produce trails from sugar baits (Henquell, 1976; Horstmann, 1976); moreover, Rosengren and Fortelius (1987) found trail recruitment to protein baits by F. grankullensis (closely allied to F. aquilonia >. The trails left by red wood-ants consist of faecal fluid, which is evacuated from the hind gut by ants with a full crop (Horstmann, 1976). Such trail-laying appears involuntary and thus distinct from the deposition of recruiting pheromones by ants of some other species (Wilson, 1962a,b; Hangartner, 1969; Hiilldobler et al., 1978; Pasteels et al., 1987; Hijlldobler and Wilson, 1990). In the present study, trail-laying and trail following will be considered “trail communication” in the sense of Rosengren and Fortelius (1987): that is, the trail left by a replete forager provides naYve foragers with directional information enabling them to find the food on which the replete had fed. Rosengren and Fortelius (1987) term this process “directional recruitment”. To the investigator, recruitment is perceptible as an increase with time in the numbers of ants arriving per unit time at a source of food. At first, after discovery of the food, this increase in traffic with time should in theory be approximately exponential: on returning from feeding, the discoverers of the food induce nestmates to travel to the source, whereupon each newcomer at the source becomes in its turn a “discoverer” and recruits one or more fresh foragers on its return to the nest. Because there is a finite supply of unoccupied foragers within the nest, the probability of a returned ant’s encountering a potential recruit will decrease with time. Hence the rate of recruitment, and consequent traffic to the food, will eventually produce a logistic curve with time. The shape of the curve of traffic against time is not, however, a guide to the presence or absence of recruitment (cf. Wenner et al., 1967). A logistic curve of traffic against time might result if the ants were responding to an odour diffusing from the food, or to the presence of increasing numbers of their nestmates at the food. Again, if the ants lay trails on the way from the nest (Brian, 19831, increasing densities of these trails might have a similar effect. Moreover, as ants discover the food independently the traffic will tend to increase, since individuals that have found the food are likely to revisit it. Alone, revisiting will cause a linear increase in traffic, coming to an upper limit as the supply of new foragers is exhausted. Superimposed on this linear increase would be any exponential increase, whether caused by recruitment or by other factors. Unless the newcomers could be distinguished from individuals which had previously visited the food, the exponential curve would probably not be discernible against the linear “background” curve. Recruitment
113
would then reveal itself simply by a greater linear increase in traffic to the food over time, relative to the “background” increase observed in conditions where recruitment did not occur. Where recruitment does occur, moreover, it may take one or more different forms: each mode of recruitment that is employed will contribute to the increase in the numbers of foragers arriving at the food with time. In the simplest case, successful foragers encountering nestmates may (i) induce them, by excitation, to leave the nest to seek food (Rosengren, 1971). Further, they may (ii) pass on directional information during contact with nestmates (Cosens and Toussaint, 1986; Reznikova and Ryabko, 1990). Finally, directional information may be exchanged without direct contact between ants, if (iii) the successful foragers leave a trail from the food. Excitation (i) will lead to an increase in the number of ants leaving the nest: some of them will find their way to the route that leads to the food. If the outgoing foragers obtain directional cues during contact (ii), they will go to this route without loss of time: thus, while both (i) and (ii> will cause an increase in traffic on the route leading to food, (ii) takes effect more quickly and will therefore lead to a steeper rise in the increase of traffic over time. Trail-laying (iii) will make this rise steeper still, and will probably increase the numbers of ants travelling to the food as well, since a trail may elicit a response from many individuals simultaneously. The present study was an attempt to demonstrate trail communication by separating its effects from those of other influences on foraging traffic. Trail communication can occur only where the successful foragers walk home over a route to which potential recruits have access. It would not occur if the ants used a homeward route different from the outward route leading to the food. An experiment was therefore designed in which the foraging traffic on a one-way route to a feeding site was compared with that on a route giving free access between a feeding site and the nest. It was predicted that, if trail recruitment occurred, it would produce a greater increase over time in the traffic on the two-way route, relative to that on a one-way route on which recruitment could occur only through contact or excitation.
Materials
and Methods
Wood-ant
colonies
Four colonies of Formica aquilonia were collected from a site at Glen Tanar, Deeside, Scotland (grid reference NJ 466 938), three in the autumn of 1989 and the fourth in the autumn of 1990, and established in the laboratory. Day-length was artificially prolonged to 14 h during the winter months, and temperatures in the laboratory ranged from 8 to 22”C, so that the ants remained active throughout the year. The ants were fed carbohydrate in the form of a 0.17 M aqueous solution of sucrose, with blowfly pupae (Caliphora sp.) or mealworms (Tenebrio sp. larvae) as prey (Lamb and Ollason, 1993). The work described in this paper was carried out in the summer and autumn of 1991.
Experimental
apparatus and procedures
Three Perspex boxes (20 x 8 x 8 cm), designated box 1, box 2, and box 3, were made available as novel feeding sites to the foragers of one colony. The boxes were open at the
114
top, but ants were prevented from climbing out by a coating of Silicone Shine spray lubricant (Hycote Ltd., Salmon Fields, Royton, Oldham, Lanes., England) on the inner walls and on the outer face on the nestward side. Because the ants will not walk on this lubricant, the floor of each box was lined with paper. The boxes (Fig. I) were placed in a row above the substratum of one of the nests. The entrance to each box was connected to the nest by a 28-cm paper bridge. A two-way bridge led to box 1, the control, which contained no food. A one-way bridge led to box 2, in which was placed a dish of 0.7 M aqueous sucrose solution; box 2 was furnished with a one-way exit. A two-way bridge led to box 3, in which there was a dish of food as in box 2. The ants entered each box through a Perspex pipe 2.5 cm in diameter, inserted in one end of the box and projecting 4 cm inside (Fig. I>. From the pipe, the ants reached the floor of the box by walking down a cardboard strip or by dropping from the pipe. The ants left box 1 and box 3 by walking back up the cardboard strip, and thence to the bridge: since the inner walls of the boxes were coated with Silicone Shine, the pipes formed the only exit from box 1 and box 3. Box 2, however, was provided with a 0.7 cm wide slot provided in the floor, under the entrance pipe. The paper lining the floor of box 2 was folded so as to project 1.5 cm downward from the slot; near the corner of this exit, a glass funnel was suspended by a paper strip which projected into the bell of the funnel. The ants readily climbed down the paper strip or the outside of the funnel, dropping the last 2-5 cm to the nest floor below. Box 2 was also furnished with a “non-return valve” in its entrance pipe. This was formed by a small glass tube with a tapered end (Quickfit 14/23), through which the ants walked when entering: each of the three boxes was fitted with a tube (Fig. 1). The ants readily walked out of the tubes through the tapered ends, but were reluctant to re-enter them. When leaving box 1 and box 3, the ants were able to walk over the outsides of the tubes. In box 2, however, the outside of the tube was coated with Silicone Shine, and because the ants avoided walking on this substance they tended to leave the box by the slot. In general, the ants, on finishing feeding, appeared to explore the nestward end of the box before they found the exit, gathering underneath the projecting pipe; hence, when leaving box 2, they found the slot at once. Before each trial, the colony to be used was deprived of sucrose for one day. On the following day, the three boxes were placed above the nest in numerical order, with the bottom ends of their bridges in contact with the nest or the substratum and not more than 8 cm apart. Before food was placed in box 2 and box 3, traffic to each box was measured by calculating the mean rate of arrival of ants per minute from ten I-min counts per bridge, taken during one 30-min counting session. During this preliminary session, the paper lining in box 2 was positioned so as to cover the slot, in order to ensure that, at the start of a trial, roughly similar numbers of ants were inside box 2 and box 3; the slot was opened before food was placed in the boxes. Dishes of sucrose solution were then placed in box 2 and box 3. Rates of arrival of ants on each bridge were recorded during four 30-min counting sessions, spaced at intervals of 30-60 min. Ten counts were made per bridge per session, thus 40 counts were made per bridge, with the baits in place, in each trial. The ants had access to the boxes for about 6 h during a trial, after which the bridges were lifted from the substratum and any ants remaining in the boxes were returned to the nest. Six trials were carried out in summer, using all four colonies. Six more trials were performed in the autumn, using the two colonies that appeared most active. Different colonies were used for consecutive trials; for each different trial using the same colony, the
115
gloss
tube
cardboard
control (Box 1)
level
entrance
of nest substrate
pipe
ff
paper bridg
food/./..tluowoy
, except for the first session in the second series (p = 0.941). Comparisons between means showed that the mean proportion of the original traffic was significantly greater on the two-way bridge than on either of the other bridges in every session
except
the
first;
a
the
one-way
and
control
bridges
b
40
35.
35
30.
30
25
25.
25.
20
20.
15 10
f
3
I
f
f
c
z
20.
15,
f
0
E
I
0
12
;
f
5
5,
3
P
z
0
12
8 .
i
0
12
o4
0
12
3
4
f P
1 P
IO.
E 3
I
15.
5
differ
i1
1
40
f.
not
c
40 -
35 30 I
did
3
4
3
4
30min courhng sess~ons unbaited wnlml one-way bridge two-way bndge e
d 40
f
35 1
f
30 25 m
I
f:
E
d
15
e f
10 5
0 . .
40
40
35
35
30
30
25
25
20
m
15
15
10
ii
5
;
E
i
Ei
0
0 I___ 0
12
3
4
10 5 0
0
12
3
4
30mm coutiing sessions unbaded control one-way bridge two-way bndge
15 . .
Fig. 3. Changes with time in the mean numbers of ants per minute on the three bridges in the second series of trials. (a) Trial 1; (b) Trial 2; (c) Trial 3; (d) Trial 4; (e) Trial 5; (0 Trial 6.
119
a 9-
a. 7. 6. 5. 4. 3.
I i
2-
ri
P
1.
b
97 a. 7. 6. 54. 3. 2.
c
l-
- . 3
2
1
0
4
30min counting sessions
““baIted
control
13
l l
one-way bridge
two-waybridge
Fig. 4. Changes with time in the rate of arrival of ants on the three bridges, as a proportion mean rates of arrival during an unbaited counting session proportion
of the original mean ants/min
of the
(see text). Open circles represent
the
for the control bridge; triangles, for the one-way bridge;
and closed circles, for the two-way bridge. (a) First series; (b) Second series.
significantly.
Differences
bridge (p < 0.001). the original highest trials
bridge
2a-f
thereafter.
Fig. 4a,b
proportion although
show time
traffic.
On all bridges,
each counting
activity shows
session,
for
on the two-way the timing
bridge,
they did so, however,
with
the highest
sessions
how the mean during The
different
traffic
the mean
means during trials.
did not increase session
of ants arriving On
series
more
significant mean
the third
some
bridge
days,
counting
of trials:
rapidly
for
significantly
during
in the second per minute the
traffic in both
series.
on each
one-way
bridge
to increase
session
of
giving the
with
and declined
on each bridge series,
the
than that on the one-way
on both bridges.
each
proportions
session
also tended
of the original
and second was similar
with
at the third
proportions
highly
bridges,
the third
numbers
highest
bridge increased
of the response
were
session,
on the control
was usually the first
(ii)
and the two-way
in the second
2). On the control
series;
and 3a-f
changed
received
counting
both the one-way
began to increase
value (Table
in the first
Figs.
time.
traffic
between
On
in
mean bridge,
120
Fig. 5. During replete summer, bridge
ants
the experiments, had
when leading
walked.
The
a trail of spots often appeared spots (presumably
the ants collected into
much
prey.
faecal
on paper over which
fluid)
were
most
often
large numbers seen during
of the
The photographshows the top end of the two-way
box 3. The trail was laid down
during
a single trial.
(The ruled
lines are 8 mm
apart.)
During the trials in summer, the papers over which replete ants walked became spotted with a brown liquid (Fig. 5). These spots were associated with routes leading away from food: the cardboard strip connecting the floor of box 3 with its entrance pipe, the bridge to box 3, and the paper lining projecting from the slot in box 2. They did not always reflect a large increase in traffic during a trial.
Discussion The experiment separates the effect of recruitment by trail-laying from other influences on the rate of arrival of ants at a source of food. If trail recruitment is excluded, an increase in traffic to the food with time might be due to one or more of these factors: an odour from the food diffusing increasingly with time; cues from ants already feeding and from trails laid during exploration; excitation of nestmates and recruitment by contact with successful foragers; and return visits by foragers which have previously fed at the site. Except for excitation, which would probably affect all three bridges, all these factors would be expected to act equally on ants approaching the one-way and two-way bridges. Hence, the significantly greater increase with time of the traffic on the two-way bridge may be attributed to trail recruitment by replete foragers returning to the nest. The idea of a trail gains support from the observation that returning repletes frequently left a line of brown spots on the substrate: the brown substance was probably gut fluid, which was shown by Horstmann (1976) to mediate orientation to food. The trail would have helped those ants which had previously fed in box 3 to relocate and return to the site: but there is no reason
121 to suppose that naive ants would not also respond to these trails.
Hence, increases with
time in the traffic on the two-way bridge were due to the arrival of relatively greater numbers of both experienced and naive foragers. It was predicted that, due to factors arising from the presence of food, traffic to both baited sites would increase more with time than would traffic to the control site. Within a session, however, there was no significant difference in the mean proportions of original traffic on the control and the one-way bridges. On the other hand, traffic to both the baited sites increased more rapidly than did traffic to the control site. Presumably, the ants were guided to the one-way bridge by cues from the food or from nestmates: odours, trails laid on the outward journey, recruitment by excitation or contact. The present work was not designed to distinguish between factors other than trail recruitment. Among ants in general, trail-laying and recruitment are likely to be experimentally detectable when the colony is hungry and food of (Wilson, 1962a; Hangartner, 1969, 1970; Hiilldobler 1985, 1986; Rosengren and Fortelius, 1987). In this starving after one day without sucrose, they were perceived the 0.7 M sucrose as a high-quality find,
relatively high quality is discovered et al., 1974; Cosens and Toussaint, study, while the ants were far from presumably hungry: they may have particularly as they usually fed on a conditions conducive to trail recruit-
weaker solution. It is possible that for F. aquilonia, ment are peculiar to the early spring, when the ants have begun to forage but before the aphids have produced offspring (Rosengren and Sundstrom, 1987; Sudd, 1987). At this time the ants rely on sap leaking from trees (Rosengren and Sundstriim, 1987), or forage on bracken, which produces nectar in early summer (Sudd and Sudd, 1985; Sudd, 1987). During the rest of the ants’ active season, into the autumn, they visit aphids of various
species as they become productive (Skinner, 1980a,b; Whittaker, 1991>, and most foragers learn where aphids may reliably be found. Many foragers become site-faithful, and thus inflexible in their choice of a feeding site (Rosengren and Sundstriim, 1987): hence, during most of the season, trail-laying from sugary foods might prove to be ineffective as well as unnecessary. On the other hand, recruitment of foragers to underexploited sites would enable the foraging population to maximise returns from the resources in its territory; for recruitment would allow the distribution of the population to approach more closely the distribution of the available honeydew. If it became distributed exactly in proportion to the distribution of food, the foraging population would conform to the ideal free distribution of Fretwell and Lucas (1970): this would maximize the colony’s rate of intake of honeydew (Lamb and Ollason, 1993). If the distribution of aphids within the territory were to vary throughout the season, any approach to the ideal free distribution would improve the foraging populations’ success over that obtained if the distribution of foragers were static. The distribution of aphids is, in fact, likely to vary during the season, both on the host tree and among trees. Aphids aggregate on a tree at different sites depending on the season, for the position of nutrient sinks (points of growth or of senescence> varies throughout the summer (Dixon, 1970; Wratten, 1974; Carter and Maslen, 1982; Klingauf, 198713). Moreover, aphids of different species feed at different sites on a tree (Carter and Maslen, 1982; Kidd et al., 1990), and different species of aphids attain their peak numbers at different times in the season (Carter and Maslen, 1982). When aphids reach peak densities on a host tree, winged forms develop, dispersing to found new colonies on other trees (Kidd and Tozer, 1984): they may fly far without alighting (Klingauf, 1987a), but on finding a high-quality feeding site they will begin to produce wingless offspring (Kawada, 1987). It is thus possible for a productive site to vanish in a few hours, while new sites may
122
quickly become densely populated because of the parthenogenetic growth of clones of aphids (Dixon, 1987). Thus, trail communication may occur throughout the summer with undiminished usefulness to the colony. The potential for trail communication is always present, for successful foragers may respond to repletion at any time by leaving trails of hind-gut fluid. Whether or not trail recruitment ensues depends upon other foragers’ willingness to follow trails: in other words, upon the presence in the foraging population of a sufficient proportion of novice foragers without site allegiance (Rosengren and Sundstrom, 1987). Because experienced foragers may be lost with equal probability throughout the season, the foraging populations numbers must be continually supplemented by newcomers, so the population should always retain some flexibility of behaviour. Through trail communication, the distribution of the foraging population may be able to adjust itself, through feedback, to changes in the distribution of sugary food. Replete ants laying trails from any rich and underexploited feeding site are likely to recruit fresh foragers to the site: the number of foragers visiting such an underexploited source of food will consequently increase, until the site is fully exploited. This point is reached when the foraging population is distributed as the ideal free distribution. In the laboratory, foragers of F. aquilonia have been found to approximate to the ideal free distribution (Lamb and Ollason, 19931, even when large numbers of foragers show site allegiances (Lamb and Ollason, in preparation). Evidently, trail communication plays a part in bringing about that distribution of the foraging population which tends to maximise the intake of energy by the colony.
Acknowledgements We are grateful to Mark Young for and to Andy Lucas for the photograph. criticisms of the first draft of this paper. to A.E.L. from the Natural Environment
advice and encouragement throughout this study, We also thank two anonymous referees for helpful This work was supported by a research studentship Research Council.
References Brian, M.V.,
1983.
Carter, Cl. HMSO,
Social Insects: Ecology and Behavioural Biology. Chapman and Hall, London.
and Maslen,
N.R.,
1982.
Conifer Lachnids in Britain.
Forestry
Commission
Bulletin
58.
London.
Cosens, D. and Toussaint, Formica aquilonia.
Cosens, D. and Toussaint, aquilonia
N., 1985.
Anim.
An experimental
study of the foraging strategy of the wood ant
Behav., 33: 541-552.
N., 1986.
The dynamic nature of the activities of the wood ant Formica
foraging to static food sources
within
a laboratory
habitat. Physiol.
Entomol.,
11:
383-395. Dixon, A.F.C., 1970. The effect of the quality of phloem sap on reproduction and body size in the sycamore aphid. In: A. Watson (Editor), Animal Populations in Relation to Their Food Resources, Blackwell, Dixon, A.F.G.,
Oxford, 1987.
and P. Harrewijn
pp. 270-287. Parthenogenetic reproduction and the rate of increase in aphids. In: A.K. Minks (Editors),
Aphids:
Crop Pests. Elsevier, Amsterdam, Fretwell,
S.D.
distribution
and Lucas, H.L.,
1970.
in birds. I. Theoretical
Their
Biology,
Natural Enemies and Control,
Vol. 2A, World
pp. 269-287. On territorial
behavior and other factors influencing
development. Acta Biotheor.,
19: 16-36.
habitat
123 D.M.,
Gordon,
Rosengren,
Ecol. Entomol., Hangartner,
W.,
Physiol.,
1969.
L., 1992.
The allocation of foragers in red wood ants.
20.
Trail
laying in the subterranean
ant, Acanthomyops
interjectus.
J. Insect
15: 1-4.
Hangartner,
W.,
1970.
Acanthomyops Henquell,
R. and Sundstrom,
17: 114-I
Control
D., 1976.
90:
Holldobler,
105-I
ant,
dans les conditions
U.,
1974.
Tandem
running
in Camponotus.
I. Comp.
27.
B., Stanton,
Novomessor
in the odor trails of the subterranean
26: 664-665.
lnsectes Sociaux, 23: 577-583.
B., Moglich, M. and Maschwitz,
Physiol.,
quantity
Sur I’existence d’une piste chimique chez Formica polyctena
de vie semi-naturelle. Holldobler,
of pheromone
Mayr. Experientia,
interjectus
R.C. and Markl,
(Formicidae,
Holldobler,
6. and Wilson,
Horstmann,
K., 1976.
H.,
Hymenoptera). E.O.,
1990.
1978.
Recruitment
and food-retrieving
behavior in
I. Chemical signals. Behav. Ecol. Sociobiol.,
4: 163-181.
The Ants. Belknap, Cambridge, MA.
Uber die Duftspur-Orientierung
bei Waldameisen
(Formica polyctena
Forster).
lnsectes Sociaux, 23: 227-242. Kawada, K.,
1987.
(Editors),
Amsterdam, Kidd,
Polymorphism
and morph
determination.
In: A.K.
Minks
and P. Harrewijn
Aphids: Their Biology, Natural Enemies and Control, Vol. 2A. World Crop Pests. Elsevier,
N.A.C.,
pp. 255-268. Smith,
chemistry
S.D.J.,
Lewis,
and the population
Hunter and N.A.C.
G.B.
and Carter,
dynamics
Kidd (Editors),
C.I.,
1990.
of conifer aphids.
Interactions
In: A.D.
between host-plant
Watt,
S.R.
Leather,
M.D.
Population Dynamics of Forest Insects, Intercept, Andover, MA,
pp. 183-193. Kidd, N.A.C.
and Tozer,
D.J., 1984.
large pine aphid, Cinara pinea. Klingauf, F.A., Aphids:
Their
Amsterdam, Klingauf,
1987a.
F.A.,
Aphids:
Exp. Applic., 35: 37-42.
Host plant finding and acceptance. In: A.K. Minks and P. Harrewijn
Biology,
Natural
Enemies
and Control,
198713. Feeding adaptation and excretion. Biology,
Natural
Enemies
World
Crop
(Editors),
Pests.
In: A.K. Minks and P. Harrewijn
and Control,
J.C., 1993.
Foraging wood-ants
Formicidae) tend to adopt the ideal free distribution. Lamb, A.E. and Ollason,
Vol.
2A,
World
Crop
Pasteels, J.M., Deneubourg, recruitment
formica
aquilonia
Elsevier, (Editors),
Pests.
Elsevier,
in foraging wood-ants
J.L. and Goss, S., 1987. Information
Sensory Systems and Communication
98. Formica aquilonia
of foragers in a regenerating environment. Self-organisation
mechanisms in ant societies (I):
to newly discovered food sources. Experientia Suppl.,
Sz. and Ryabko, B., 1990.
Yarrow (Hymenoptera:
Behav. Proc., 28: 189-l
J.C. (in preparation) Site fidelity
Yarrow and its influence on the distribution
Reznikova,
2A,
pp. 225-253.
Lamb, A.E. and Ollason,
Trail
Vol.
pp. 209-223.
Their
Amsterdam,
Host plant and crowding effects in the induction of alatae in the Entomol.
54: 155-l
75.
theory approach to communication
in Arthropods.
in ants. In:
Advances in Life Sciences. Birkhauser,
Basel,
pp. 305-307. Rosengren,
R., 1971.
Route fidelity, visual memory and recruitment
of the genus Formica (Hymenoptera, Rosengren,
R., 1977.
Acta Zool.
Foraging strategy of wood ants (Formica
topographic traditions. Rosengren,
Formicidae).
Acta Zool.
R. and Fortelius,
W.,
Fenn.,
1986.
149:
R. and Fortelius,
W.,
red wood ants (Formica). Rosengren,
R. and Sundstriim,
1987.
Ann. Zool. L., 1987.
133:
rufa group).
I-1
06.
I. Age polyethism
and
l-30.
Ortstreue
in foraging ants of the Formica
hierarchy of orienting cues and long-term memory. Rosengren,
behaviour in foraging wood ants
Fenn.,
rufa group -
lnsectes Sociaux, 33: 306-337.
Trail communication
and directional recruitment
to food in
Fenn., 24: 137-146. The foraging system of a red wood ant colony (Formica
s.str.1 - collecting and defending food through an extended phenotype. Exp. Supp., 54: 117-l 37. Rosengren, R. and Sundstriim, L., 1991. The interaction between red wood-ants, Cinara aphids, and pines. A ghost of mutualism Oxford
University
past? In: C.R. Huxley
Press, Oxford,
pp. 80-91.
and D.F. Cutler (Editors),
Ant-Plant
Interactions,
124 Skinner,
G.J., 1980a.
(Hymenoptera:
Territory,
Formicidae)
trail structure in limestone
and activity patterns in the wood ant, Formica woodland
in north-west
England. J. Anim.
rufa
Ecol., 49:
381-394. Skinner,
C.J., 1980b.
The feeding habits of the wood ant, Formica
in limestone woodland in north-west Sudd, J.H.,
1987.
Wenner,
A.M., Wells, J.B., 1991.
(Editors), Wilson, Wilson,
1985.
Seasonal changes in the response of wood-ants to sucrose baits.
P.H. and Rohlf, F.J., 1967.
1962a.
Zool.,
An analysis of the waggle dance and recruitment in
40: 317-344.
Effects of ants on temperate woodland trees. In: C.R. Huxley Interactions,
Oxford
Communication
University
1962b.
Communication
II, An information SD.,
1974.
Press, Oxford,
among workers
I. The organisation of mass-foraging. Anim.
E.O.,
Smith). Wratten,
Ant-Plant
E.O.,
Smith).
Formicidae)
10: 89997.
honey bees. Physiol. Whittaker,
rufa (Hymenoptera:
Ecol., 49: 417-433.
individual behaviour and mixed diet strategy in ants. Exp. Supp., 54: 81-92.
Sudd, J.H. and Sudd, M.E., Ecol. Entomol.,
England. J. Anim.
of the fire ant Solenopsis
Behav., 10: 134-l
among workers
saevissima
(Fr.
saevissima
(Fr.
47.
of the fire ant Solenopsis
analysis of the odour trail. Anim. Behav., 10: 148-158.
Aggregation in the birch aphid Euceraphis
food quality. J. Anim.
and D.F. Cutler
pp. 67-79.
Ecol., 43:
191-I
98.
punctipennis
(Zett.)
in relation to