Sulcopolistes atrimandibularis, Social Parasite and Predator of an Alpine Polistes (Hymenoptera, Vespidae)1

Ethology 86, 71-78 (1990) 0 1990 Paul Parey Scientific Publishers, Berlin and Hamburg ISSN 0179-1613

Dipartimento di Biologia Animale e Genetica Universita di Firenze and Dipartimento di Morfofisiologia Veterinaria Universitd di Torino

Sulcopolistes a trimandibularis, Social Parasite and Predator of an Alpine Polistes (Hymenoptera, Vespidae)' RITACERVO,MARIACRISTINA LORENZI & STEFANO TURILLAZZI CERVO,R., LORENZI,M. C. & TURILLAZZI, S. 1990: Sulcopolistes atrimandibularis, social parasite and predator of an Alpine Polistes (Hymenoptera: Vespidae). Ethology 86, 71-78.

Abstract Sulcopolistes atrimandibularis Zimmermann is the obligate social parasite of Polistes biglumis bimaculatus Geoffry in Furcroy, a mountain species of paper wasp. Unlike all the other hymenopteran social parasites, the Sulcopolistes female obtains part of the food for her.immature brood by plundering other nests of the host species. Parasite females can control more than one host nest: one of them (the nursery nest) she uses solely for reproduction purposes (from which the Sulcopolistes reproductives emerge) and others (supply nests) are used for exploiting the Polistes brood content. It is possible that this behaviour is an adaptation to the extremely short colonial cycle of the host. Dipartimento di Biologia Animale e Genetica, Corresponding author: Dr. Stefan0 TURILLAZZI, via Romana 17, 1-50125 Firenze.

Introduction Obligate and permanent social parasites in wasps can be defined as those

with n o worker caste and whose females have lost the capacity for nest building and rearing their own reproductives, relying entirely on the host species for this purpose (TAYLOR 1939; cf. WILSON1971). In the Polistinae, the only known social parasites which fit this definition are the three species belonging to the genus Sulcopolistes which live around the Mediterranean Sea. Very little information is available on the biology and social behaviour of these species, some authors preferring to include them under the genus Polistes, together with the species they parasitize (WILSON1971; CARPENTER 1990). The most comprehensive study available is due to SCHEVEN (1958) who reported on all three species. Sulcopolistes

' Study supported with funds of the Minister0 della Pubblica Istruzione (60%). U.S. Copyright Clearance Center Code Statement:

0179-161 3/90/8601-0071$02.50/0

72

RITACERVO. MARIACRISTINA LORENZI & STEFANOTURILLAZZI

atrimandibularis parasitizes Polistes biglumis bimaculatus, a mountain species (DE BEAUMONT & MATTHEY1945; SCHEVEN 1958; CERVO et al. 1990 a), although it has also been found on nests of Polistes gallicus L. (formerly P. foederatus Kohl of authors; DAY1979; SCHEVEN 1958; pers. obs.). A P. biglumis bimaculatus nest invaded by this parasite is easily recognizable in more advanced stages of development because of the pale colour of the opercula of the parasite which differ strikingly from the dark ones of the Polistes. We recently had the opportunity to continue studies on these parasites and to observe the behaviour of female Sulcopolistes atrirnandibularis in the field and in the laboratory. This paper describes the relationships of the parasite with a population of Polistes biglumis bimaculatus in the western Alps.

Materials and Methods 20 colonies of P.b. bimuculutus parasitized by S. atrimundibuluris (17 parasites) were found and catalogued in the summer of 1987, 32 in 1988 (23 parasites) and 36 in 1989 (26 parasites). All colonies came from a south facing alpine meadow (186C-2000 m a d ) near Montgenivre (Dep. of Hautes Alpes, France) a few km from the French-Italian border. Nests were checked for size (no. of cells) and condition of immature brood on their discovery and almost every week afterwards. At every check the number of individuals present on the nest was recorded, and any newly emerged wasps were individually marked with enamel paints. Observations in 1987, 1988, and 1989 totalled 32, 34 and 42 h. In particular, 6, 14 and 12-h observations over the three years focused o n parasites which visited two o r more host colonies. Observations were made in the hours of the day when the wasps were most active (from 9.30 to 17.30 h), and when meteorological conditions were suitable (sunny days with light to moderate wind) from 24th JuI. to 1st Sep. (for a total of 13 d) in 1987, from 10th Jul. to 27th Aug. (13 d) in 1988 and from 14th Jun. to 20th Aug. (22 d) in 1989.

Results A. Primary and Secondary Nests

Three of 17 parasites in 1987, 5 of 23 in 1988, and 5 of 26 in 1989 were observed to spend part of their time on two nests or more. We defined the nest where the parasite stayed longer as the “primary nest”, and any other nest where the same marked parasite was found as a “secondary nest”. In 1987, we observed that the 3 parasites spent an average of about 60 % of the study time on the primary nests and 21 % on the secondary nests; we could not tell where they spent the remaining time. In 1988, the time the 5 parasites stayed on the primary nest averaged 63 % of total time, on the secondary nests 17 % and elsewhere the remaining 20 %. In 1989, the 5 parasites stayed on the primary nests for an average of 53 % of the time, 17 % on secondary nests and the remaining 30 % elsewhere. Visits (n = 101) to the secondary nests by the parasites (n = 13) lasted, on average, 4 min (range 20 s-15 min). During the observation period the parasites (n = 13), on average, left the primary nests every 17 min (range 6 m i n 4 4 min, calculated on 173 exits). We never observed two parasites on the same nest (primary or secondary) at any time.

73

A Social Parasite and Predator of an Alpine Polistes

B. Characteristics of Parasitized Nests

The characteristics of the secondary nests on their discovery in the field and those of the respective primary nests at the same time are given in Table 1. Table I: Characteristics of primary (p) and secondary (s) nests usurped by Sulcopolistes utrimandibuluris on discovery in the field (-: no data) Opercula pale P S

broken P S

Adults of host P S

0 0 0

0 0 0 0 0 0

4 2 2 2 4 1

20 13

0 0

25 0 23

0 0 0 0

0 0 0 2 0 0 5 0 2 0 3

4 0 3 2 1 6 3 3 2 0 2

0 1 8 8 25 4 4 29 16

0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0

1 3 5 5 5 2 1 4 0

Date

Distance between p and s nest

Cells P S

Eggs P S

Larvae P S

dark P S

1987 2517 2517 1718

1.0 m 7.0 m 5.0 m

57 38 59 39 57 46

24 6 18 2 2 13

1 7 9 2 5 5 1 6 6

4 9 4

1 0 0

0

0 6

16 11 2 2 5 0 1 0 0 9 4 2 0

0 0

4 1 0 3 0 3

- 10 3 - 1 - 0 0 - 1 0 0 7 6 0 0 1 1 1

1988 1418 1418 16/8 19/8 6/9

42.5 5.8 2.0 0.8 7.5 0.2

m m m m

53 43 52 30 41 57 42 43 30 74 94

1989 117 13/7 2617 2617 11/8 4/8 818 1518 18/8 20/8

2.0 m 26.0 m 3.5 m 15.0 m 9.0 m 21.0 m 13.0 m 10.0 m 21.0 m 24.5 m

- 43 32 - 3 5 - 5 3 69 34 - 2 5 56 50 50 54 69 32 71 50

m m

2 1 0

0 0

0

0 2 0 7 0

- 12 12 - -

-

0

0 0 0 3 0 0

15 4 1 0

16 4 2 2 10 5 3

14 0 0 2 11 0 4

0 7 0

2 2 0 0 5 9 7 0 1

1 1 1 1 1 1 3 2 5

The distance between the primary and secondary nests ranged from 20 cm to 42.5 m (Table 1). While usurpations of primary nests are known to occur, at the earliest, between 15th of Jun. and the beginning of Ju1. when the nests were still in the preemergence period (CERVO et al. 1990 a), secondary nests were only found later: on 25th Jul. and 17th Aug., 1987, on several days from 14th Aug. to 6th Sep., 1988, and from 1st Ju1 to 18th Aug., 1989 (Table 1). The actual dates of secondary nest invasion can only be given for colonies periodically monitored before event, and are available only for 4 parasites in 1989 (between 1st Jul. to 15th Aug.) and 1 in 1988, which was observed to attack a secondary nest in Sep. The most striking difference between the primary and secondary nests, which is apparent from Table I, is the total absence of pale opercula (Sulcopolistes pupae) in the secondary nests, but which were found in 13 out of 17 primary nests

74

RITA CERVO,

MARIACRISTINA LORENZI&

STEFAN0

TURILLAZZI

(2 nests subsequently showed pale opercula, and 2 nests were raided later). When the secondary nests were discovered in the field, they were significantly smaller (U-test; p < .OOI) and contained fewer immature brood (U-test; p < .OOI) than their associated primary nests. There was also a larger number of host females (workers and gynes, see LORENZI & TURILLAZZI 1986) on the primary nests than on their respective secondary ones (U-test; p < .O1). Broken opercula had holes in their caps, and the cells were usually empty or contained damaged pupae. Dark, broken opercula were only present on secondary nests. Three other nests showed some intermediate characteristics falling between those of the primary and secondary nests: the parasites (which did not appear to have a secondary nest) spent much of their time there, but they were small in size, and the cells were closed with dark opercula. All these nests, however, had one pale operculum by the end of the season (end of Aug.). C . Destruction of Brood in Primary and Secondary Nests

Six parasites (out of the 13 that had secondary nests and which were observed for a total of 32 h) were seen on 11 occasions extracting larvae and, after breaking the opercula, pupae from the cells of secondary nests. They then chewed them and took them back to the primary nest, where they offered the bolus to the adult Polistes and the larvae. In 108 observation hours on primary nests (for which we were unable to find secondary nests) 8 different parasites were observed on 20 occasions bringing food back to their own nests. As we never saw the parasites hunting for other prey in the field and it was usually possible to recognize parts of Polistes larvae and pupae in the loads they brought back to the primary nests, they may well have had one or more secondary nests in the area. Brood destruction on the primary nest, on the contrary, was only observed three times during 108 observation hours: twice by Sulcopolistes (on larvae of unknown species) and once by a Polistes (on Sulcopolistes pupa). Brood destruction, evident from the periodical nest checks, allowed a more complete comparison of treatment of primary and secondary nests. In 1988 we registered a total of 22 partly eaten Polistes pupae (dark opercula with a small hole) and 9 missing larvae (of unknown species) in the 6 secondary nests. O n the 23 primary nests, 9 larvae (of unknown species) and 9 pupae of Sulcopolistes (pale opercula) had been destroyed; 7 of these latter pupae were found killed after the disappearance of the parasite female. D. Productivity in Primary and Secondary Nests O u r data can only give an idea of nest productivity. This is due to the fact that any individuals that emerged from a nest and disappeared between one check and another could not be counted. In addition, in the secondary nest it was impossible to know how many pupae actually emerged or were preyed upon by the parasite. However, from the colour of the operculated cells in the combs, it is obvious that, from a certain moment onwards, only reproductives of the parasite emerge from the primary nests, while none did from the secondary nests. O n the

A Social Parasite and Predator of an Alpine Polistes

75

primary nests, which only bore pale opercula from about mid-August onwards, emergence of host Polistes was hampered by the parasites (LORENZI et al. subm.). The fall in the number of Polistes offspring seen in primary nests is due to the destruction of host immature brood by the parasite immediately after invasion (CERVOet al. 1990 a), as well as to her control over reproduction capacity of the host (CERVO 1990). Destruction of host pupae and larvae on the secondary nests drastically reduced host productivity. O n 15th Aug. 1 9 8 8 , l l primary nests had an average of 53 (SD = -I- 10) cells compared to an average of 38 (SD = -I- 12) cells in 20 randomly collected unparasitized nests from the same area (which therefore probably included both non parasitized and secondary nests) (U-test; p < .O1). O n 9-14th Aug. 1989, 6 primary nests had an average of 52 (SD = 2 17) cells, not statistically different (U-test; ns) from the average of 42 cells (SD = k 7) of 11 nests found in the same area which were definitely not parasitized. E. The Parasite Behaviour on Primary and Secondary Nests The first and most frequent action the parasite females performed upon landing on the secondary nests was dominating the hosts (in 49 out of 94 cases, 52 %) and cell inspection (69 out of 153 cases, 45 YO) on the primary nest. Upon landing, the dominance activity of the parasites against any P. b. birnaculatus females present was significantly higher on a secondary than on a primary nest (p < .005; Wilcoxon test). Fig. 1 gives the average percent distributions of the activity of parasite females on all primary and secondary nests. Cell inspection and grooming are practised more on the secondary nests, while rest and ventilation occur more on the primary nests (differences between pairs of a behaviour of a certain parasite on her primary and secondary nest pooled over the whole observation time, are statistically significant; Wilcoxon test, see Fig. 1). Aggression and dominance behaviour are performed more against Polistes females on secondary nests than 6o

=

1

prlmory nests

0setondory nests

*

n

Fig. I:

++

Average distributions of the nest activities of parasite females on the primary and secondary nests (:>:>p < ,005, '> p < ,025; Wilcoxon test)

76

RITACERVO, MARIA CRISTINA LORENZI& STEFANOTURILLAZZI

against those on primary nests (p < .005;Wilcoxon test). Polistes females reacted differently to parasite aggression depending on whether they belonged to the primary or secondary nests: after 112 assaults observed on secondary nests, in 80 cases host females left within 30 s (71 %), while on the primary nests this happened in only 10 cases out of 45 (33 %) (p < .005;Wilcoxon test). The host females on the primary nests, after dominance and aggressive interactions with the parasite, performed various nest activities. However, in all of the parasite-host interactions the Sulcopolistes female proved to be dominant over her host whether on the primary or the secondary nest. Once on primary nest, the parasite assumes the role of alpha female within the very first hours of her arrival (CERVO et al. 1990 a).

Discussion Two kinds of relationships exist between S. atrimandibularis and its host

P.b. bimaculatus: social parasitism and predation. In actual fact this species does not fit the definition of obligate social parasite given by TAYLOR (1939) and behaves differently from the social parasites of vespine wasps (cf. AKRE1982, for a review of the matter) and, as far as we know, from the other Sulcopolistes (SCHEVEN 1958; pers. obs.). To rear her brood, a S. atrimandibularis female has to rely on the work of her host as well as her own, although no quantification could be made at this stage of the study. Part of the food for her own immature brood comes from raiding other nests of the host species, where she always behaves aggressively. In P. b. bimaculatus the number of workers per nest is extremely low as future gynes are present from the first stages of the post-emergence period (LORENZI & TURILLAZZI 1986), so the parasite has a very low work force at her disposal. In addition, as brood rearing in Sulcopolistes begins about 3 4 weeks later than in Polistes, the parasite breeding season is very short. The parasite female is incapable of founding a nest; we never observed any construction activity, only two cases in which the parasite enlarged an already existing cell. With this exception, her behaviour is very similar to that of her host foundress who, perhaps because of the extremely short brood rearing season, continues her external activity (pulp collection and foraging for immature brood) even in the post-emergence period (LORENZI & TURILLAZZI 1986). Thus the parasite exploits two kinds of nest: the primary nest which she uses solely for reproduction purposes (“nursery nest”), where her reproductives emerge, and one or more secondary nests which she plunders and which function exclusively as “supply nests”. The predation of larvae and pupae in Polistes has occasionally been described in P. dominulus Christ (PARDI1951), in P. jadwigae Dalla Torre and in P. chinensis antennalis PCrez (SAKAGAMI & FUKUSHIMA 1957; KASUYA et al. 1980), but in these cases predation is only intraspecific, contrary to what we observed in

S. atrimandibularis. The behaviour of the parasite on the two kinds of nests (Fig. 1) also reveals their different function. O n the primary nest, she stays longer but is often inactive. O n the contrary, when she arrives on a secondary nest, the Sulcopolistes

A Social Parasite and Predator of an Alpine Polistes

77

female first dominates the Polistes and then chases them away. During her short stay there she is very busy controlling the cells and plundering their contents. SCHEVEN (1958) had already distinguished two kinds of parasitized nests; some were large and filled with the immature brood of the parasite, others were small and empty. In all probability they corresponded to the nursery and supply nests. The invasion time of the supply nests was only established for five parasites (on 10 secondary nests), but nevertheless some considerations can still be made. The parasite may usurp two or more nests at the beginning of the season, and then choose one as her nursery, leaving the others as supply nests (as we actually observed in one case). In addition, the parasite may usurp another supply nest when the previous one has been completely exploited. One parasite, for example, was observed plundering 6 supply nests one after the other in the summer of 1989, but she may well have used more than six nests over the whole season. Exploitation of the supply nests expecially affects the host pupae; the highest food demands from the Sulcopolistes larvae (highest number of developed parasite larvae on the nursery nest) roughly coincide with the highest number of host pupae in the supply nests. Although we cannot be sure whether we observed the very first visit to a secondary nest, the behaviour of S. atrimandibularis on supply nests suggests she invades them more violently than she does a primary nest (CERVOet al. 1990 b). SCHEVEN (1958) observed aggressive usurpation when he experimentally introduced parasite females in cages where post-emergence colonies of P. b. bimaculatus with several workers were reared. We believe that the author re-created the situation of a secondary nest invasion or at least he simulated the invasion strategy used by the parasite when the season is fairly advanced. By predating on pupae the parasites can exploit the supply nests for a longer period, while the larvae destined to be future victims are still developing. This phenomenon leads to intense exploitation of the Polistes population. It is probable that in 1988 the parasitized primary nests were larger than randomly collected unparasitized Polistes nests from the same area because some of the latter had already been exploited in part by the Sulcopolistes, even though they did not show any signs of predation. The periodic checks of P. b. bimaculatus nests showed that in August 1989, 8 out 47 controlled colonies (17 Yo) had definitely been plundered by parasites; this percentage increases to 28 % if the colonies (n = 5) with broken opercula are also considered, but where a Sulcopolistes female was never seen. Another 11 % of the colonies (5 out 47) were represented by primary nests. The need for several secondary nests for each Sulcopolistes could also imply severe competition between individual parasites. In this respect, we think we can give an explanation for the colour of the parasite opercula which are strikingly different from the dark ones of the host. It has been assumed that the dark colour of the opercula is a way of favouring warming the nest and accelerating the development of the immature brood (LORENZI & TURILLAZZI 1986); however it is possible that the pale opercula of the parasite acts as a signal towards conspecifics. An already parasitized nest could be plundered by some other parasite female, especially if the “owner” is absent. This could lead to fights between the parasites,

78

CERVO,LORENZI & TURILLAZZJ, A Social Parasite and Predator of an Alpine Polistes

with possible great harm to both contenders. A parasite female can easily recognize a nest with sparkling pale opercula, and thus avoid risking landing there. This hypothesis, of course, waits to be tested, but broken white opercula have never been found (Table 1). The kind of social parasitism in S. atvimandibulavis seems to differ from true inquilinism in which the only activity of the parasite queen is to lay eggs and control (by both behavioural and pheromonal mechanisms) the host queen and workers to induce them to rear her own brood. The relationship between S. atrimandibuluvis and its host, although at first sight quite complex, could be due to the parasite adapting to the extremely short colonial cycle of the Polistes so allowing her to raise more brood in the season. The evolutionary history of this behaviour and the phylogenetic relationships between this parasite and her host and with other species of Sulcopolistes are interesting questions for future research. Acknowledgements We wish to thank Prof. L. PARD]who read and criticized the manuscript and C. COSTER LONGMAN who helped with the English text.

Literature Cited AKRE,R. D. 1982: Social wasps. In: Social Insects, Vol. IV. (HERMANN, H. R., ed.) Acad. Press, N e w York, pp. 1-105. J. DE & MATHEY, R. 1945: Observations sur les Polistes parasites de la Suisse. Bull. SOC. BEAUMONT, vaudoise Sci. nat. 62, 439-455. CARPENTER, J. M. 1990: Phylogenic relationships and the origin of social behaviour in the Vespidae. R. & Ross, R., eds.) Cornell Univ. Press, in In: The Biology of Social Wasps. (MATTHEWS, press. CERVO,R. 1990: I1 parassitismo sociale nei Polistes (Hymenoptera, Vespidae). Ph.D. Thesis, Univ. of Florence, Florence _ _ , LORENZI,M. C . & TURILLAZZI, S. 1990a: N o n aggressive usurpation of the nest of Polistes biglumis bimaculatus by the social parasite Sulcopolistes atrimundibularis (Hymenoptera, Vespidae). Insect. soc., in press. _ _ ,_ - & - - 1990b: O n the strategies of host nest invasion in three species of Sulcopolistes, social parasite of Polistes wasps. Actes Insect. SOC.,in press. DAY,M. 1979: The species of Hymenoptera described by Linnaeus in the genera Sphex, Chrysis, Vespa, Apis and Mutillu. Biol. J. Linn. SOC.12, 45-84. Y. & ITO, Y. 1980: O n “intercolonial ” cannibalism in Japanese paper wasps, KASUYA,E., HIBINO, Polistes chinensis antennalis Perez and P.jadwigae Dalla Torre (Hymenoptera: Vespidae). Res. POPUI.EcoI. 22, 255-262. S. 1986: Behavioural and ecological adaptations to the high mountain LORENZI,M. C . & TURILLAZZI, environment of Polistes biglumis bimaculatus. Ecol. Entomol. 11, 199-204. PARDI,L. 1951 : Studio della attivith e della divisione di lavoro in una societh di Polistes gallicus (L.) dopo la comparsa delle operaie. Arch. 2001. ital. 36, 363-431. S. F. & FUKUSHIMA, K. 1957: Reciprocal thieving found in Polistes fudwigae Dalla Torre. SAKAGAMI, J. Kans. Entomol. SOC.30, 140. SCHEVEN,J . 1958: Beitrag zur Biologie der Schmarotzerfeldwespcn Sulcopolistes atrimandibularis Zimm., 5. semenowi F. Morawitz und S. sulcifer Zimm. Inscctes SOC.54, 4 0 9 4 3 7 . TAYLOR, L. H. 1939: Observations on social parasitism in the genus Vespula Thomson. Ann. Entomol. SOC.Am. 32, 304-315. WILSON,E. 0. 1971 : The Insect Societies. Belknap Press, Harvard Univ., Cambridge.

Received: March 7, 1990 Accepted: July 10, 1990 (W. Pflumm)