© Entomologica Fennica. 15 March 2007
Behavioural and socially parasitic relations between Polyergus rufescens (Latr.) and Formica polyctena Först. (Hymenoptera: Formicidae)
Wojciech Czechowski
Czechowski, W. 2007: Behavioural and socially parasitic relations between Polyergus rufescens (Latr.)and Formica polyctena Först. (Hymenoptera: Formi- cidae). Entomol. Fennica 18: 5464. Polyergus rufescens (Latr.), an obligate slave-maker, and Formica polyctena Först., an aggressive, territorial wood ant species, rarely co-occur in the field, and there are almost no data on their mutual relations under natural conditions. These interactions were studied in the Bia³owie¿a Forest (NE Poland), based on two P. rufescens colonies (with Formica fusca L. slaves)nesting within the territories of F. polyctena. The wood ants routinely searched the immediate vicinity of P. rufescens nests, whereas P. rufescens ants raided F. fusca colonies very close to F. polyctena nests or their columns passed right next to them, they eventually crossed wood ants foraging and removal routes, and even directly attacked F. polyctena colonies and robbed their brood. Interspecific relations in these partic- ular situations are described and discussed in the contexts of supposed chemical camouflage/mimicry of P. rufescens and interspecific competition hierarchy in ants. W. Czechowski, Museum and Institute of Zoology PAS, Laboratory of Social and Myrmecophilous Insects, Wilcza St 64, PL-00-679 Warszawa, Poland; E-mail: [email protected] Received 19 May 2006, accepted 11 December 2006
1. Introduction reviews see Czechowski 1996, 2003b)and be - havioural ecology (e.g. Punttila et al. 1996, Cze- Formica sanguinea Latr., the European blood- chowski 2000, Czechowski & Vepsäläinen 2001, red ant, and Polyergus rufescens (Latr.), the Eu- Czechowski & Markó 2006). P. rufescens is a ropean Amazon ant, are Palaearctic, facultative rare oligotope of dry grasslands, and therefore it and obligate respectively, slave-making social has little chance of encounters with wood ants in parasites of Serviformica For. ant species. F. the field. No wonder that reports on relations be- sanguinea, a common polytope of dry habitats tween P. rufescens and wood ants under natural (including forests), frequently co-occurs with conditions are scarce. Thus Forel (1900)de - wood ants (Formica s. str.)and, unavoidably, of - scribed a queenright (polygynous)colony of F. ten comes into direct contact with them; it is even pratensis Retz. (otherwise a polytope of dry habi- able to use wood ants as (atypical)slave species. tats)mixed with P. rufescens and F. (Servi- There is abundant literature regarding these inter- formica) fusca L. workers. As for behavioural actions, both with regard to social parasitism (for data, Dobrzañski & Dobrzañska (1978)and ENTOMOL. FENNICA Vol. 18 Relations between P. rufescens & F. polyctena 55
Dobrzañska & Dobrzañski (1989)mentioned a P. wood ant F. polyctena. PR-I inhabited a ne- rufescens colony that was gradually destroyed as glected coniferous-deciduous forest plantation, a result of heavy losses when its raids crossed a nesting in a part of it which was not intensively route of nearby F. rufa L. More recently, Cze- searched by wood ants coming from a vast multi- chowski (2006)reported that a F. polyctena nest polydomous system consisting mainly of Först. route was a factor indirectly promoting small, often unstable stump-side nests; its raids, supposed emancipation of F. fusca slaves from however, entered the intensively searched part of colonies of P. rufescens. Mabelis (2000)noticed the habitat (see Czechowski & Markó 2006 and that P. rufescens (with F. fusca slaves)did not re - Fig. 1 in Czechowski 2007). PR-II lay in a dry act aggressively towards F. rufa workers in the light spruce-pine forest, within the area inten- vicinity of its own nest, and concluded that the sively searched by F. polyctena, at a point where latter might also be enslaved. Deconinck et al. wood ants routes forked, 7.5 m away from the (2001)subsequently announced that conclusion nearest fairly big F. polyctena nest FP-I (Æ»60 as a fact, most probably too rashly. Then, as so far cm), a member of a well established polydomous there has been no reliable confirmation of P. colony comprising a few nests. Contacts between rufescens using Formica s. str. workers as atypi- the two species were observed in the following cal slaves. However, on the analogy of F. sangui- situations: (1)everyday encounters of P. rufes- nea, another slave-maker, this possibility can not cens (and its F. fusca nestmates)with F. polyc- be ruled out. tena in the vicinity of the Amazons nests, (2) The aim of this paper is to present the results crossings of F. polyctena nest-areas or its forag- of field observations of spontaneous relations be- ing routes by P. rufescens raiding columns, and tween the Amazon ant P. rufescens and F. po- (3)direct raiding of P. rufescens on F. polyctena lyctena, a common wood ant species. These rela- colonies. tions included mutual behavioural responses of the heterospecific individuals (and especially their groups)encountering in various situations, 3. Results as well as socially parasitic relationship between the slave-maker and F. polyctena as a possible 3.1. Relations in the vicinity atypical host species. of the P. rufescens nests
P. rufescens nest PR-I lay in the zone only 2. Study area, material and methods sparsely searched by wood ants. However, in 2002, that nest was investigated by the nearest The investigations were conducted in the (10.2 m) F. polyctena colony, which nested in a Bia³owie¿a Forest (NE Poland)near the village of shady, overgrown place and was evidently look- Topi³o from 2001 to 2003 from mid-July till mid- ing for a new nesting place (establishing small August, i.e. within the period of P. rufescens raid- short-lived trial nests here and there). Usually in ing activity. Two big colonies of P. rufescens, the afternoons, single F. polyctena scouts came PR-I (found in 2001), nesting inside a rotten right up to nest PR-I, thoroughly searched the stump and in the soil at it, and PR-II (found in area and returned straight to their nest. During the 2002), nesting in a low earth bank, were observed reconnaissance they avoided contact with occa- practically every day within the study periods sionally encountered Amazon workers and their when the weather seemed to be good enough for numerous F. fusca slaves; the former paid no at- Amazons raids. Their raiding columns were esti- tention to these scouts and the latter sometimes matedtocompriseupto2,000workers.Thecal- harmlessly chased the intruders over a few centi- culations (very rough estimates)were based on metres (compare to the event described in subsec- the number of individuals crossing an imaginary tion 3.3). line in a unit of time, and the time it took the entire The mean dynamic density (± standard devia- tight column to cross that line. tion)of dispersed wood ant workers around the Both colonies nested within territories of the other P. rufescens nest, PR-II, measured at noon 56 Czechowski ENTOMOL. FENNICA Vol. 18 ca. 1 m away from the main nest holes, was 15.5 ± Markó 2006). The raided nest itself was situated 5.2 inds. × 0.25 m2 ×5min1 (n = 6). Besides their 4.5 m away from one of those two nests. That frequent encounters with enslaved F. fusca, Amazons raid did not provoke any response which were a matter of course in those circum- from nearby colonies of F. polyctena or even visi- stances, there were not even the slightest conflicts ble interest on the part of wood ant workers which between them and no aggressive behaviour to- were densely searching the area. Later both out- wards single Amazon ants was recorded. During bound and inbound raiding columns of PR-I were mass raiding activities of P. rufescens (recruit- seen many times to pass very close to nests of F. ment circling around the nest, marching out and polyctena (even less than 0.5 m from mound return of raiding columns), dispersed F. edges), again with no or hardly any response of polyctena usually simply moved aside without the wood ants. Once a group of about 20 F. po- protest. Aggressive responses of some F. polyc- lyctena workers coming down the mound at- tena towards P. rufescens were rare. Such aggres- tempted to attack an inbound column of P. ru- sive wood ant individuals were generally driven fescens but their attack was prevented within sec- away or sometimes even immediately killed. The onds by a few Amazon ants from the column only violent conflict noticed close to nest PR-II which launched a counter-strike without letting happened in 2002, when a pre-raid circle of ex- go of the pupae in their mandibles. tremely excited Amazon workers, so-called ac- Foraging routes of F. polyctena within the tivists according to Dobrzañskas and Dobrzañ- raiding distance of colony PR-I were not very dis- skis (1960)terminology, moved too close to the tinct. P. rufescens columns would cross them main F. polyctena FP-I foraging route that ran 1 without the slightest difficulty and frightened mfromtheP. rufescens nest entrances. That sec- wood ant foragers would then scatter towards the tion of the route was 1.5 m wide and the mean dy- sides or wait on grass blades for the column to namic density (± SD)of wood ants within it was pass. as high as 766 ± 42 inds. × 0.25 m2 ×5min1 (n = Raids of colony PR-I on F. fusca co-nesting 5). Only this route, and only in this (initial) sec- with F. polyctena in one stump (a plesiobiosis fre- tion, was impassable for P. rufescens, both single quently observed in this habitat; see Czechowski scouts and raiding columns (see below), and nor- 2002)took place three times during one season mally the Amazons did not try to move in that di- (2002). Even though the closest nest holes of the rection. On the day of the conflict, the outermost two plesiobiotic nests were not more than 20 cm P. rufescens individuals were fiercely attacked by apart, neither the attacking Amazons took interest the wood ants and about a dozen of them each in the nearby wood ants nests (or the ants them- assailed by tens of infuriated F. polyctena work- selves), nor the latter reacted to the conflict taking ers were killed with no response from the rest of place in their immediate vicinity. A similar situa- their swarm (see also Czechowski 2006 and Fig. tion was also observed when the F. fusca colony 2 in Czechowski 2007). raided by P. rufescens PR-I formed a plesiobiotic relationship (also in a stump)with Tetramorium caespitum (L.). Again, the close neighbour did 3.2. Relations on the routes not react in any way. On another occasion an Am- of P. rufescens raidingcolumns azon column walked over a soil mound of T. caespitum on their way, again without any re- The event that marked the starting point for ob- sponse of the latter. The raiding columns of P. servations of the relations between the Amazons rufescens PR-II also walked unpunished over and the wood ants was a successful 55-meter raid nests of Lasius niger (L.)and L. platythorax Sei- of P. rufescens colony PR-I on a F. fusca nest sit- fert, and on one occasion these Amazons entered uated within the area intensively searched by F. and penetrated a nest of Myrmica sp. without any polyctena. Along the way towards their goal (and harm to either party. back)the raiding column passed between two P. rufescens colony PR-II nested and oper- small F. polyctena nest situated only 6.5 m apart ated within a territory much more numerously (FP-IV and FP-V; see Fig. 4 in Czechowski & populated by F. polyctena in comparison with ENTOMOL. FENNICA Vol. 18 Relations between P. rufescens & F. polyctena 57 that settled by PR-I and so PR-II faced much column, without mishap or even slowing down, larger wood ant colonies, resulting in more spec- entered and crossed a dense trail of wood ants, tacular relations between Amazons and wood which parted in front of the former like the bibli- ants during such encounters as described above. cal Red Sea. That raid turned out to be a failure; Colony PR-II also freely raided F. fusca near one meter behind the trail the head of the column nests of F. polyctena and the latters foraging stopped and, after searching unsuccessfully for routes, and PR-II columns would cross wood their target (beyond the F. polyctena trail)for a ants routes practically unpunished. Thus, a raid few minutes the Amazons began to go back. The on a target nest situated only 1 m away from a disruption of F. polyctena traffic lasted about 10 small (Æ = 40 cm)but very active F. polyctena min. During that time wood ants gathering on nest provoked only slight excitement among both sides of the column captured and killed sev- wood ants. The inbound and outbound route of eral (not more than 20)individuals of P. rufescens another raid ran along the edge of a sand bank sur- without any response or interest from the part of rounding the much larger mound of nest FP-I (see their nestmates. above), but the response of its inhabitants was The other occasion, several days later, in- similarly restrained. volved a P. rufescens outbound column ap- Two days before the fierce conflict between proaching the removal trail of F. polyctena at a F. polyctena workers from their foraging route sharp angle. At the point when it met the trail, the towards P. rufescens circlers at nest PR-II de- column stopped and the Amazons started to circle scribed in subsection 3.1., a raiding column looking for their target. The search circle of P. crossed the same wood ants route at a point about rufescens ants spread all across the trail, forcing 10 m away from the place of that conflict. The wood ants aside. The trail was blocked for about route was 1 m wide at that point and the mean dy- 15 minutes until the Amazon ants found their tar- namic density of wood ant workers within it was get, a F. fusca nest situated a meter away from 165 ± 17 inds. × 0.25 m2 ×5min1 (n = 5). Since there. This time, the aggressive response of F. po- the Amazons crossed the route at a sharp angle, lyctena was even weaker, with very few victims the raid and the foraging route overlapped over a (three P. rufescens workers were seen to be distance of about 1.5 m and wood ants traffic was killed). At the same time some wood ants carry- only slightly disturbed. No incidents were no- ing larvae and pupae sought refuge on herbs to ticed during the crossing of the Amazon columns wait for the commotion to end, with the Amazons (both outbound and inbound)with the wood ants paying no attention to them just like before. route. An especially unusual sight was that of P. A totally conflict-free crossing of an Amazon rufescens carrying pupae, walking in loose array raiding column (both outbound and inbound)was hand in hand with wood ant foragers carrying also observed for colony PR-I, which walked their prey. Even single straggling P. rufescens en- across a removal trail (less busy than the one de- tered the column of F. polyctena workers without scribed above)of a F. polyctena colony raided(!) hesitation and were able to get to the other side a few days earlier (see below). without difficulty. In the next season (2003) F. polyctena colony FP-I formed a filial nest approximately 25 m 3.3. Direct raids of P. rufescens away from the maternal nest. Throughout the pe- on the wood ants nests riod of observations both nests were connected by a very busy 3040 cm wide two-way removal P. rufescens colony PR-I was the only one to in- trail, on which wood ants used to carry cease- vade nests of F. polyctena. A total of eight such lessly huge numbers of larvae, pupae and adults attacks were observed in 2001 and 2002, some of (workers and fertile queens)to the new nest. The them won and some lost, some successful and trail ran 4 m away from P. rufescens nest PR-II. some failed. The targets were situated 3.38.2 m Two raids of Amazon ants collided with that trail. away from nest PR-I (see Fig. 1 in Czechowski On the first occasion, the outbound column met 2007). Most were small unstable colonies of F. the removal trail at a right angle. The head of the polyctena, with several hundred ants in each that 58 Czechowski ENTOMOL. FENNICA Vol. 18 moved from place to place possibly due to over- fusion, came back to their nest. It was only during growing (shading)of the habitat. These raids of their third attempt that they finally managed to P. rufescens on wood ants nest can be divided reach the nest of F. fusca which was probably into accidental and intended (i.e. originally di- their original target. rected to that very, previously detected, target)on Of the six raids that could be regarded as in- the basis of the circumstances and behaviour of tentional attacks on wood ants colonies, includ- the Amazon ants. ing four successful attempts, the most spectacular Two of the raids were definitely accidental. raid took place in 2001. The target was a F. po- On one occasion, in 2001, a column was halfway lyctena nest situated 6 m away from PR-I. The through (3.3 m away from PR-I)to the F. fusca mound had a diameter of 20 cm and was 8 cm nest when it bumped into a very small (incipi- high. The moment the Amazons launched an at- ent?) F. polyctena colony in a clump of grass tack, queens (three were seen)and workers with filled with fine plant nest material. A part of the pupae began to flee the nest immediately. Some column attacked the nest while the rest continued climbed on nearby plants but most of them dis- their march. Wood ant workers immediately be- persed in all directions. The nest was attacked di- gan to flee up plant stems, some carrying pupae rectly by a relatively small part of the Amazon (they still had not come down two hours after the forces, but this group of attackers was still numer- raid). No fighting was observed but the smell of ous enough to cover the surface of the mound formic acid could be felt in the air. The Amazons completely. The remaining Amazons surrounded began to carry away pupae and larvae, gaining a the nest, forming a dense ring of 0.5 m width. The spoil of about 200. When the main column of P. Amazon ants forming the cordon were extremely rufescens was returning empty-handed from their excited, paying no attention to F. polyctena actual target near (or actually over)the nest being workers with pupae that were trying to break looted, again only a very small part joined in the through the cordon to get away from the danger action and the rest simply continued their march. zone (this is a characteristic behaviour of P. ru- However, on the next day, P. rufescens carried fescens, observed also when it is confronting its out another raid on the same F. polyctena nest, typical host species; Dobrzañska & Dobrzañski this time deliberate and involving the entire col- 1988). Only after all wood ants had left the nest umn, and gained several tens more pupae and lar- did the surrounding Amazons enter it to carry vae. away several hundred pupae and larvae that were The other accidental raid, in 2002, was left there. As soon as P. rufescens workers for- against a temporary bivouac of a small F. polyc- med the home-bound column when leaving the tena colony migrating from dense shrubs. The attacked nest, F. polyctena that were dispersed bivouac had been unluckily established on a line around the nest returned to it in massive numbers. connecting nest PR-I (4.1 m away from it)with a Now they covered the surface of the mound F. fusca nest nearby. When the head of the col- densely and attacked Amazon stragglers leaving umn encountered the bivouac, wood ant workers the nest with their spoils. Victims were few: not with pupae dispersed all around. Some Amazon more than a dozen or so F. polyctena workers ants also scattered and returned to their nest while killed at the moment of attack and an equal num- the rest, despite some defensive attempts that ber of P. rufescens stragglers. At least one dealate lasted a few seconds, took over the bivouac, gain- gyne of P. rufescens took part in the raid, but it re- ing about 100 pupae. Then F. fusca slaves that turned to the maternal nest with the inbound col- had been following the Amazon column searched umn. the area for a few hours, collecting abandoned pu- F. fusca slaves also took part in that raid of P. pae (at least several tens were found)and injured rufescens on F. polyctena. They followed the or dead wood ants, including a half-dead F. Amazon column to the site of the attack, mingling polyctena queen. The Amazons carried out a sec- with some of its rear, less numerous segment. ond raid in the same direction immediately after They did not attack the F. polyctena nest, but returning to their nest, but stopped at the site on went on to collect pupae dropped by the fleeing the (former)bivouac and, after a moment of con - wood ants. Later, when all P. rufescens had left ENTOMOL. FENNICA Vol. 18 Relations between P. rufescens & F. polyctena 59 the site, F. fusca searched for pupae hidden in the the latter with its enslaved pupae of F. (Servi- grass for a few more hours. Notably, enslaved F. formica) cinerea Mayr and F. fusca, a total of fusca workers would never do this during raids on about 3,000 individuals were put out 2 m away nests of their conspecifics, but those targets were from nest PR-II. F. fusca slaves immediately usually situated much farther away (up to 55 m in started to pick them up, assisted by massive num- the case of PR-I)from the P. rufescens nest (ex- bers of P. rufescens workers forming a very busy cluding special integration raids; see Czechowski two-way route between their nest and the pupae 2005). pile. Next year, however, no slaves other than F. Two of the raids on the wood ants met defeat. fusca were found in colony PR-II. In one case, in 2001, only the final part was ob- served: the P. rufescens column was retreating, there was a distinct smell of formic acid in the air, 4. Discussion andontheF. polyctena mound wood ants were killing a few captured Amazons. A later defeated The observed encounters between F. polyctena raid in 2002 was followed from the beginning to and P. rufescens in the field were usually charac- the end. The target was not, as usual, a very small terized by mutual indifference of these ant spe- migrating colony but a new trial nest of a larger cies. On most occasions, the response of wood colony of F. polyctena (>1,000 individuals) the ants to the presence of Amazon ants, if there was one whose scouts had explored and would later any, could perhaps best be described as the for- explore P. rufescens nest PR-I (see subsection mer being intrigued by the unusual situation or 3.1). The trial nest, situated under a log, was 8.2 m abnormally high concentration of strangers rather away from PR-I. It was raided by one of two col- than assuming an aggressive attitude. F. polyc- umns that the original outbound column had split tena would only display aggression (usually into a few meters away from PR-I. The raiding quickly turning into panic)when facing a direct arm consisted of no more than 1,000 workers. attack of P. rufescens on its nests, but in such situ- The attack was very violent and so was the ensu- ations aggressive behaviour is obvious. The only ing defence and counter-attack of the wood ants. occasion when F. polyctena displayed overt ag- Following a few minutes of fighting all the Ama- gression towards P. rufescens outside their own zon forces began to flee at the same moment, run- nest was when wood ants from a foraging route ning away in panic in all directions. After a while attacked careless Amazon circlers which had most began to retreat to their nest in disarray, but come too close to the route (see Czechowski tens of scattered Amazons were still wandering 2006). It appears that this aggression should be around the area or could be found high up on associated with enormous (pheromone-induced) blades of grasses and herbs a few hours later. Sev- agitation of Amazon workers in that particular eral dozen were killed on the F. polyctena nest in pre-raid phase, comparable to the level of agita- the first phase of the conflict. Three days later P. tion while attacking their target nest. This is also rufescens raided the same target, this time using related to the fact that only the part of the F. all its forces, only to find the trial nest completely polyctena foraging route, adjoining the P. ru- abandoned; after the first attack the F. polyctena fescens recruitment arena, was impassable for the colony had returned to its own permanent nest. Amazon ants. This is a good illustration of a point Both in 2001 and 2002 colony PR-I of P. made long ago by J. Dobrzañska and J. Dobrzañ- rufescens gained at least 200300 worker pupae ski, who wrote that ants are able to distinguish a and last instar larvae of F. polyctena (that was the passive attitude from a combative attitude of number observed but raiding periods of Amazon other species, and their response actually depends ants lasted longer than the observations). How- on the behaviour of the latter (Dobrzañska & ever, no enslaved wood ant workers were found Dobrzañski 1962, Dobrzañska 1976). ever in PR-I later. As a footnote, it should be men- A particularly spectacular aspect of the inter- tioned that in 2002 an experiment was carried out specific relations described in this paper was that where a pile of worker pupae of Formica Amazons raiding columns were able to freely (Coptoformica) exsecta Nyl. and F. sanguinea, cross, almost without losses, the dense and wide 60 Czechowski ENTOMOL. FENNICA Vol. 18 routes of wood ants. It was very surprising, espe- cinerea, F. rufibarbis and F. lusatica, unlike F. cially in the light of previous reports of the entire fusca (and F. cunicularia Latr., another frequent P. rufescens colony dying out under similar con- slave species), can be very aggressive towards ditions (Dobrzañski & Dobrzañska 1978 and other ant species, both in socially parasitic and Dobrzañska & Dobrzañski 1989; unfortunately competitive relations. They are known for their these authors provide no information on the traf- fierce (and often effective)defence of their nests fic intensity on the F. rufa route crossed by P. against slave-makers (Dobrzañska 1978a, Do- rufescens or on the size of the F. rufa colony). It is brzañska & Dobrzañski 1989, Grasso et al. 1994, possible that the distinctly different attitudes of Le Moli et al. 1994b, Mori et al. 1994, Czechow- wood ants towards P. rufescens in apparently the ski & Rotkiewicz 1997, Seifert 1997, DEttore et same situation were motivated by the presence of al. 2004, Czechowski & Radchenko 2006, A. different slave species in those Amazon colonies, Mori, pers. comm., W. Czechowski, unpubl.). a consequence of which were their different col- Formica cinerea, in respect of its status in in- ony labels. Despite some uncertainties in this re- terspecific competition hierarchy, is potentially a gard (see Liu et al. 2003), slave-maker ants be- territorial species (Markó & Czechowski 2004), come olfactorily similar to the slave species pres- able to successfully compete for territory against ent in the colony. In Polyergus species this is wood ants (Dlussky 1967, Czechowski & Markó achieved through chemical camouflage (Yama- 2005 and unpubl.). Big colonies of the two re- oka 1990)or chemical mimicry (e.g. Habersetzer maining aggressive Serviformica species, espe- & Bonavita-Cougourdan 1993, Bonavita-Cou- cially those of F. lusatica, are also supposed to be gourdan et al. 1996, 1997, DEttore & Heinze potentially territorial (Seifert 1996). Interspecific 2001). Thus, slave-makers which co-exist with competition hierarchy assumes that nesting of different slave species can be perceived variously one territorial species on the territory of another by other ants, including both their hetero- and territorial species is not possible, and the occur- conspecifics. The above hypothesis is supported, rence of intruders invariably ignites a conflict. At for example, by enormous intraspecific aggres- the same time, F. fusca is a non-aggressive, non- siveness in F. sanguinea when a confronted party territorial, subordinate submissive species which includes a large proportion of F. polyctena slaves represents the lowest level in interspecific domi- (Czechowski 1994). If this assumption is true, nance hierarchy. An individualistic and opportu- then the attitude of F. polyctena towards P. nistic manner of foraging and evasive behaviour rufescens should generally resemble its attitude enables F. fusca not only to nest fairly close to towards free living ants of a given slave species, wood ant colonies but also to search freely which is, in turn, related to the position of the lat- (though not without limitations)right next to their ter in interspecific competition dominance hierar- mounds (see Savolainen 1990, 1991). According chy (see Savolainen & Vepsäläinen 1988, Pisar- to the hypothesis proposed above, F. polyctena ski & Vepsäläinen 1989). could have perceived Amazons nesting within Like all wood ants, F. polyctena is a typical their territories or crossing through as F. fusca- territorial species, the frequent top dominant of like ants that have a natural right to stay there. forest ant assemblages (Savolainen & Vepsä- Further studies are needed to clarify the proposed läinen 1988, 1989, Savolainen et al. 1989). The mechanism. papers of Dobrzañski & Dobrzañska (1978)and During a conflict Amazons are never the first Dobrzañska and Dobrzañski (1989)that describe to start individual physical combats; they react, the destruction of a P. rufescens colony by wood using their mandibles and, rarely, formic acid, ants do not mention its slave species. Other stud- only when they have been attacked themselves ies by the same authors indicate that their P. (e.g. Czechowski 1975b, Mori et al. 2001). Help- rufescens enslaved F. cinerea and F. rufibarbis F. ing a nestmate whose life is in danger is also not as well as F. fusca. It is now known that the name characteristic of this species (Czechowski 1977, F. rufibarbis could also refer to F. lusatica Seifert 2003a)(unlike F. sanguinea; see e.g. Czechow- (see Seifert 1997)(= F. glauca Ruzs. sensu Seifert ski et al. 2002a). This strategy allows them to 1996 and Czechowski et al. 2002b). Formica minimize their own losses during interspecific ENTOMOL. FENNICA Vol. 18 Relations between P. rufescens & F. polyctena 61 conflicts, very spectacular examples of which brood, which is also a selection procedure. It has could be observed during their encounters with been demonstrated that slaves in P. lucidus Mayr foraging routes of F. polyctena, when there was destroy a greater proportion of heterospecific direct contact between dense columns of two po- than of homospecific slave pupae (Goodloe & tentially very aggressive ant species. The imme- Topoff 1987). J. Dobrzañski (unpubl.) came to diate retreat of an entire column, when the oppo- the same conclusion earlier on the basis of his nent is overwhelmingly strong, as is often the own observations of a P. rufescens colony. Still, case if the opponent is a wood ant colony, should Amazon colonies, at least in P. rufescens,con- also be considered an element of this strategy, as taining slaves of more than one species do occur well as a sign of some behavioural plasticity of in nature, though rarely (Dobrzañska & Dobrzañ- Amazon ants. Dobrzañska (1978b)was the first ski 1960, Dobrzañska 1976, Czechowski 1975a). to regard the escape reaction as an atavistic sign This is much more common in F. sanguinea colo- of plasticity of P. rufescens in the context of its nies (including wood ants as slaves; e.g. Cze- functional degeneration. chowski 1989), but the proportion of slaves is in- Old myrmecological literature contains some comparably smaller there (for review see Cze- information of socially parasitic relations be- chowski 1996)than in Amazons colonies, where tween P. rufescens and Formica s. str. species. it reaches 8090% (Seifert 1996). However, these are descriptions of experimental, Besides raiding wood ant nests, another inter- sometimes successful, attempts to force wood ant esting aspect of the biology of mixed colonies of workers to adopt a queen of P. rufescens (see e.g. P. rufescens (and of Amazon ants in general)and Regnier 1930). Evidence for spontaneous adop- their slaves observed in the present study was a tionofthiskindinthefieldislacking(whichdoes certain participation of slaves in some such raids. not mean that it is impossible). Forels (1900) re- P. rufescens slaves do not accompany their port, cited in the Introduction, of a natural mixed keepers in their raids as a rule (Mori et al. 1991, colony of F. pratensis/F. fusca/P. rufescens con- Le Moli et al. 1994a), this being unlike the behav- taining F. pratensis queens, actually indicates the iour of blood red ants slaves of different species opposite: the taking over of a (queenless?)Ama- (F. fusca, F. cinerea, F. cunicularia),whichmay zon colony by queen(s)of F. pratensis,asithap- happen to take part, sometimes very actively, in pens in the relations of the latter with F. san- raids (Czechowski 1998 and unpubl., Mori et al. guinea (Czechowski 2001). For the possibility of 2000, Mori et al. 2001). Forms of co-operation mutual temporary social parasitism of Formica s. between slaves and P. rufescens workers during str. species and F. sanguinea see Czechowski raiding activity of the latter known from previ- (2003b). ously published papers (Dobrzañska 1976, The present paper provides the first direct evi- 1978a)included F. fusca slaves widening nest dence of the very dulotic relations between P. holes for the inbound column carrying their spoils rufescens and the wood ants, although they in- or F. cinerea slaves taking over pupae from Ama- volved only nest raiding and robbery of F. zons on the outskirts of the nest. polyctena pupae; non-typical slaves, in all prob- However, the instances of co-operation de- ability, did not emerge. However, that does not scribed in this paper should not be interpreted as a mean that wood ant workers cannot be used as contribution of slaves to the raiding activity of slaves by P. rufescens. Most probably, it is F. their slave makers, but rather as organized activ- fusca slaves already present in the colony that are ity of F. fusca workers themselves causally and in to blame for killing wood ant workers in the terms of time related to actions undertaken by P. bud. rufescens. The departure of a raiding column al- In slave-maker ant colonies, including those ways causes agitation among slaves. They of Amazons, many of the pupae retrieved by the emerge on the surface in greater numbers and are raiders are consumed (Dobrzañski 1965, very excited, with some running a few metres Czechowski 1975a, Cool-Kwait & Topoff 1984). within and behind the column (Czechowski Since Polyergus individuals are fed by their 1975b, Dobrzañska 1976, 1978a). Reported raids slaves, it is slaves who consume the captured of P. rufescens on F. polyctena were over very 62 Czechowski ENTOMOL. FENNICA Vol. 18 short distances so that some F. fusca workers Ann. Zool. (Warsaw)43: 103126. could easily run as far as the target nest and find Czechowski, W. 1994: Impact of atypical slaves on intra- specific relations in Formica sanguinea Latr. (Hyme- scattered alien pupae. In these situations, F. fusca noptera, Formicidae). Bull. Acad. Pol. Sci., Biol. promptly starts collecting the pupae (=food Sci. 42: 345350. items), enrolling the help of more nestmates from Czechowski, W. 1996: Colonies of hybrids and mixed col- the nest. Such raiding-like, or actually foraging, onies; interspecific nest takeover in wood ants (Hyme- activity can be easily elicited in free-living F. noptera, Formicidae). Memorabilia Zool. 50, 116 + xx pp. fusca colonies (W. Czechowski, unpubl.). Never- Czechowski, W. 1998: Raids of Formica sanguinea Latr. theless, the fact remains that such behaviour of as a factor conducive to colony founding by Formica slaves, irrespective of its origin and mechanism, truncorum Fabr. (Hymenoptera, Formicidae). was an effective complementary action assisting Ann.Upper Siles. Mus., Entomol. (Bytom)8/9: 153 the activity of P. rufescens workers as the latter 157. Czechowski, W. 2000: Interference of territorial ant spe- do not pick up abandoned pupae found around cies in the course of raids of Formica sanguinea Latr. raided nests (Dobrzañska & Dobrzañski 1988). (Hymenoptera, Formicidae). Ann. Zool. (Warsaw) 50: 3538. Acknowledgements. This paper is dedicated to Assist. Prof. Czechowski, W. 2001: Mixed colony of Formica pra- Dr. Jan Dobrzañski, a Polish ethologist and myrmecolo- tensis Retz. + Formica cinerea Mayr + Formica san- gist, an eminent researcher of the slavery in ants, whose guinea Latr. (Hymenoptera, Formicidae)and its pre - 85th birthday falls in 2007. The author wishes to thank three sumed origin. Ann. Zool. (Warsaw)51: 205209. unknown reviewers, whose questions, doubts and sugges- Czechowski, W. 2002: Plesiobiosis between Formica tions helped to improve the original version of the manu- fusca L. and wood ants (Hymenoptera, Formicidae)as script. The ms benefited greatly also from an additional po- a sign of nest competition. Przegl. Zool. 46: 8186. lite review by Bálint Markó. The field study was finan- [In Polish]. cially supported by the Museum and Institute of Zoology Czechowski, W. 2003a: Polyergus rufescens (Latr.)(Hy- PAS, Warsaw. menoptera, Formicidae)and an ant lion an etholog- ical observation and its evolutionary background. Przegl. Zool. 47: 109113. [In Polish]. Czechowski, W. 2003b: Wood ants and Formica san- References guinea Latr. (Hymenoptera, Formicidae) a case of mutual social parasitism. Przegl. Zool. 47: 149 Bonavita-Cougourdan, A., Bagnères, A.-G., Provost, E., 160. [In Polish]. Dusticier, G. & Clèment, J.-L. 1997: Plasticity of the Czechowski, W. 2005: Integration raids in the Amazon ant cuticular hydrocarbon profile of the slave-making ant Polyergus rufescens (Hymenoptera, Formicidae). Polyergus rufescens depending on the social environ- Insect. Soc. 52: 103104. ment. Comp. Biochem. Physiol. B 116: 287302. Czechowski, W. 2006: The route of Formica polyctena Bonavita-Cougourdan, A., Rivière, G., Provost, E., Bag- Först. as a factor promoting emancipation of Formica nères, A.-G., Roux, M., Dusticier, G. & Clèment, J.-L. fusca L. slaves from colonies of Polyergus rufescens 1996: Selective adaptation of the cuticular hydrocar- (Latr.)(Hymenoptera: Formicidae). Pol. J. Ecol. bon profiles of the slave-making ants Polyergus 54: 159162. rufescens Latr. and their Formica rufibarbis Fab. and Czechowski, W. 2007: Do Polyergus rufescens (Latr.)and F. cunicularia Latr. slaves. Comp. Biochem. Phy- Formica sanguinea Latr. (Hymenoptera: Formicidae) siol. B 113: 313329. share resources of slave species with each other? Cool-Kwait, E. & Topoff, H. 1984: Raid organization and Pol. J. Ecol. 55. (In press). behavioral development in the slave-making ant Czechowski, W., Godziñska, E. J. & Koz³owski, M. W. Polyergus lucidus Mayr. Insect. Soc. 31: 361374. 2002a: Rescue behaviour shown by workers of For- Czechowski, W. 1975a: Mixed polycalic colony of For- mica sanguinea Latr., F. fusca L. and F. cinerea Mayr mica (Serviformica) cinerea Mayr and Polyergus (Hymenoptera: Formicidae)in response to their rufescens Latr. (Hymenoptera, Formicidae). Ann. nestmates caught by an ant lion larva. Ann. Zool. Zool. (Warsaw)33: 6775. (Warsaw)52: 423431. Czechowski, W. 1975b: Slave-raids of the ant Polyergus Czechowski, W. & Markó, B. 2005: Competition between rufescens Latr. (Hymenoptera: Formicidae). Formica cinerea Mayr (Hymenoptera: Formicidae) Przegl. Zool. 19: 449463. [In Polish]. and co-occurring ant species, with special reference to Czechowski, W. 1977: Recruitment signals and raids in Formica rufa L.: direct and indirect interferences. slave-maker ants. Ann. Zool. (Warsaw)34: 126. Pol. J. Ecol. 53: 467489. Czechowski, W. 1989: Functioning of a mixed colony of Czechowski, W. & Markó, B. 2006: Uncomfortable pro- Formica sanguinea Latr. + F. polyctena Foerst. (Hy- tection: Formica polyctena Först. defends Formica menoptera, Formicidae)with a surplus of slaves. fusca L. against Formica sanguinea Latr. (Hyme- ENTOMOL. FENNICA Vol. 18 Relations between P. rufescens & F. polyctena 63
noptera: Formicidae). Ann. Zool. (Warsaw) 56: ants. Acta Biol. Exp. (Warsaw)25: 5971. 539548. Dobrzañski, J. & Dobrzañska, J. 1978: Some questions re- Czechowski, W. & Radchenko, A. 2006: Formica lusatica lated to mechanisms of slave-raids in amazon-ant Seifert, 1997 (Hymenoptera: Formicidae), an ant spe- Polyergus rufescens Latr. Acta Neurobiol. Exp. cies new to Finland, with notes on its biology and the (Warsaw)38: 353359. description of males. Myrm. Nachr. 8: 257262. Forel, A., 1900: Fourmilière triple naturelle. Mitt. Czechowski, W., Radchenko, A. & Czechowska, W. Schweiz. Entomol. Ges. 10: 280282. 2002b: The ants (Hymenoptera, Formicidae)of Po - Goodloe, L. P. & Topoff, H. 1987: Pupa acceptance by land. Museum and Institute of Zoology PAS, War- slaves of the socially parasitic ant Polyergus.Psy- saw. 200 pp. che 94: 293302. Czechowski, W. & Rotkiewicz, W. 1997: Relations be- Grasso, D. A., Mori, A., DEttore, P. & Le Moli, F. 1994: tween Formica sanguinea Latr. and Formica cinerea Intraspecific raids and territoriality in Polyergus cinerea Mayr (Hymenoptera, Formicidae) an un - rufescens Latr. (Hymenoptera Formicidae). Ethol. usual form of dulosis. Ann. Zool. (Warsaw)47: Ecol. Evol., Special Issue 3: 8187. 469478. Habersetzer, C. & Bonavita-Cougourdan, A. 1993: Cutic- Czechowski, W. & Vepsäläinen, K. 2001: Formica rufa L. ular spectra in the slave-making ant Polyergus rufes- protects indirectly F. fusca L. against raids of F. cens and the slave species Formica rufibarbis.Phy- sanguinea Latr. (Hymenoptera: Formicidae). Ann. siol. Entomol. 18: 160166. Zool. (Warsaw)51: 267273. Le Moli, F., Grasso, D. A., Mori, A. & Ugolini, A. 1994a: Dekoninck, W., Versteirt, V. & Grootaert, P. 2001: Redis- Eco-ethological factors affecting the scouting and covery of a colony Polyergus rufescens (Latreille, raiding behaviour of the slave-making ant, Polyergus 1798)in Belgium: observations at the Hageven Na - rufescens Latr. (Hymenoptera, Formicidae). Ethol- ture reserve (Hymenoptera Formicidae). Bull. Soc. ogy 96: 289302. Royale Belge Entomol. 137: 98101. Le Moli, F., Mori, A. & Grasso, D. A. 1994b: Behavioural DEttore P. & Heinze, J. 2001: Sociobiology of slave-mak- ecology of the obligatory slave-making ant, Polyergus ing ants. Acta Ethol. 3: 6782. rufescens Latr. (Hymenoptera, Formicidae). A re- DEttore, P., Brunne, E., Wenseleers, T. & Heinze. 2004: view. Memorabilia Zool. 48: 133146. Knowing your enemies: seasonal dynamics of host- Liu, Z., Bagnères, A.-G., Yamane, S., Wang, Q. & Ko- social parasite recognition. Naturwissenschaften jima, J. 2003: Cuticular hydrocarbons in workers of 91: 594597. the slave-making ant Polyergus samurai and its slave, Dlussky, G. M. 1967: [Ants of the genus Formica (Hyme- Formica japonica (Hymenoptera: Formicidae). noptera, Formicidae)]. Nauka, Moskva. 236 pp. [In Entomol. Sci. 6: 125133. Russian]. Mabelis, B., 2000: Amazonemier (Polyergus rufescens) Dobrzañska, J. 1976: (Studies on social parasitism in ants duikt op in de Achterhoek (Hymenoptera: Formi- from the point of view of interspecific adaptative be- cidae). Entomol. Berichten (Amst.) 60: 5052 [In haviour). Kosmos A 25: 223242. [In Polish]. Dutch]. Dobrzañska, J. 1978a: Problem of behavioral plasticity in Markó, B. & Czechowski, W. 2004: Lasius psammophilus slave-making Amazon-ant Polyergus rufescens Latr. Seifert and Formica cinerea Mayr (Hymenoptera: and its slave-ants Formica fusca L. and Formica Formicidae)on sand dunes: conflicts and coexistence. cinerea Mayr. Acta Neurobiol. Exp. (Warsaw)38: Ann. Zool. (Warsaw)54: 365378. 113132. Mori, A., Grasso, D. A. & Le Moli, F. 1991: Eco-ethologi- Dobrzañska, J. 1978b: Evaluation of functional degenera- cal study on raiding behaviour of the European ama- tion of the Amazon-ant Polyergus rufescens Latr. un- zon ant, Polyergus rufescens Latr. (Hymenoptera: der an influence of socially parasitic way of life. Formicidae). Ethology 88: 4662. Acta Neurobiol. Exp. (Warsaw)38: 133138. Mori, A., Grasso, D. A., DEttore, P. & Le Moli, F. 1994: Dobrzañska, J. & Dobrzañski, J. 1960: Quelques nouvelles Specificity in host choice by the slave-making ant remarques sur léthologie de Polyergus rufescens Polyergus rufescens Latr. (Hymenoptera Formicidae). Latr. (Hymenoptère,Formicidae).Insect.Soc.7: Ethol. Ecol. Evol., Special Issue 3: 8993. 18. Mori, A., Grasso, D. A. & Le Moli, F. 2000: Raiding and Dobrzañska, J. & Dobrzañski, J. 1962: Quelques observa- foraging behavior of the blood-red ant, Formica tions sur les luttes entre différentes espèce fourmis. sanguinea Latr. (Hymenoptera, Formicidae). J. In- Acta Biol. Exp. (Warsaw)22: 269277 sect Behav. 13: 421438. Dobrzañska, J. & Dobrzañski, J. 1988: (Social parasitism Mori, A, Grasso, D. A., Visicchio, R. & Le Moli, F. 2001: in ants. Obligatory parasitism). Kosmos 37: 617 Comparison of reproductive strategies and raiding be- 639. [In Polish]. haviour in facultative and obligatory slave-making Dobrzañska, J. & Dobrzañski, J. 1989: Controversies on ants: the case of Formica sanguinea and Polyergus the subject of slave-raids in Amazon ants (genus rufescens. Insect. Soc. 48: 302314. Polyergus). Acta Neurobiol. Exp. (Warsaw) 49: Pisarski, B. & Vepsäläinen, K. 1989: Competition hierar- 367379. chies in ant communities (Hymenoptera, Formicidae). Dobrzañski, J. 1965: Genesis of social parasitism among Ann. Zool. (Warsaw)42: 321329. 64 Czechowski ENTOMOL. FENNICA Vol. 18
Punttila, P., Haila, Y. & Tukia H. 1996: Ant communities Savolainen, R. & Vepsäläinen, K. 1989: Niche differentia- in taiga clearcuts: habitat effects and species interac- tion of ant species within territories of the wood ant tions. Ecography 19: 1628. Formica polyctena. Oikos 56: 316. Regnier, A. 1930: Un cas typique dadoption: Polyergus Savolainen, R., Vepsäläinen, K. & Wuorenrinne, H. 1989: queen and rufa workers. Biol. Zentr. 50: 2531. Ant assemblages in the taiga biome: testing the role of Savolainen, R., 1990: Colony success of the submissive territorial wood ants. Oecologia 81: 481486. ant Formica fusca within territories of the dominant Seifert, B. 1996: Ameisen: beobachten, bestimmen. Formica polyctena. Ecol. Entomol. 15: 7985. Naturbuch-Verlag, Augsburg, 352 pp. Savolainen, R. 1991: Interference by wood ant influences Seifert, B. 1997: Formica lusatica n. sp. a sympatric size selection and retrieval rate of prey by Formica sibling species of Formica cunicularia and Formica fusca. Beh. Ecol. Sociobiol. 28: 17. rufibarbis (Hymenoptera, Formicidae). Abh. Ber. Savolainen, R. & Vepsäläinen, K. 1988: A competition hi- Naturkundemus. Görlitz 69: 316. erarchy among boreal ants: impact on resource parti- Yamaoka, R. 1990: Chemical approach to understanding tioning and community structure. Oikos 51: 135 interactions among organisms. Physiol. Ecol. Jpn. 155. 27: 3152.