SEARCH

SEARCH BY CITATION

Keywords:

  • chemical mimicry;
  • cuticular hydrocarbons;
  • social parasites

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Caterpillars of the parasitic lycaenid butterfly are often adopted by host ants. It has been proposed that this adoption occurs because the caterpillars mimic the cuticular hydrocarbons of the host ant. This study aimed to examine whether caterpillars of the Japanese lycaenid butterfly Niphanda fusca induce adoption by mimicking their host ant Camponotus japonicus. Behavioral observations conducted in the laboratory showed that most second-instar caterpillars were not adopted, whereas most third-instar caterpillars were successfully adopted by host workers. A chemical comparison detected no characteristic differences in the cuticular hydrocarbon profiles between second- and third-instar caterpillars. However, morphological features of the caterpillars differed between the second and third instars; third-instar caterpillars developed exocrine glands (ant organs) such as tentacle organs and a dorsal nectary organ. These results suggest that multiple chemical signatures, not only cuticular hydrocarbons, may be important for invasion of the host ant nest.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

More than 10 000 species of invertebrates have evolved as social parasites of ants (parasites that exploit the resources of an ant society), and specific adaptations enable species to infiltrate ant societies (Thomas et al. 2005). To gain access to the host society, parasites often exploit the host's chemical, vibrational and physical communication systems. Chemical mimicry (active biosynthesis of host chemicals) and camouflage (acquisition of host chemicals) of cuticular hydrocarbons are used by various social parasites including parasitic ants, crickets, silverfish and butterflies (Akino 2008), because cuticular hydrocarbon profiles are known to play a critical role in ant nestmate recognition (Martin et al. 2008; Martin & Drijfhout 2009). Typical examples are seen in the lycaenid butterflies, whose caterpillars are adopted by host ants and who either prey on host ant larvae or feed primarily on regurgitations of the ant workers. In large blue butterflies of Maculinea (Lepidoptera: Lycaenidae), chemical and acoustic signals are used to integrate with the host Myrmica ants (Akino et al. 1999; Barbero et al. 2009b). It has been shown that the caterpillars are adopted by the host Myrmica ants as a result of mimicking the cuticular hydrocarbons of the ant brood (Akino et al. 1999; Schlick-Steiner et al. 2004; Nash et al. 2008). Because a better match between the surface chemicals of Maculinea and its host species enables the parasites more easily to exploit ant colonies, accurate mimicry is proposed to be the result of a co-evolutionary arms race played out between the parasite and host (Als et al. 2001; Nash et al. 2008). However, the adoption process varies among the parasitic lycaenid species (Pierce et al. 2002; Thomas et al. 2005), and therefore the chemical strategies to be adopted will vary among species according to the pre-adoption interactions between a butterfly and its host ants. For example, on leaving the host plant, species whose caterpillars are fed by host ants (“cuckoo” species, M. rebeli and M. alcon) are quickly adopted by the worker that finds them. In contrast, species whose caterpillars prey on host ants (predaceous species, M. arion and M. teleius) await discovery by a Myrmica worker within 1–2 cm of their food plant, and an hour-long interaction between the caterpillars and host Myrmica workers is needed to establish the adoption event. It is suggested that other manipulative signals, rather than the mimetic cuticular hydrocarbons, would be involved in the adoption event of the predaceous species (Thomas 2002). However, much of the research on chemical strategies in Lycaenidae has focused on the cuckoo Maculinea butterflies.

Niphanda fusca (Lepidoptera: Lycaenidae) develops as a social parasite inside the ant colony, much like the Maculinea butterfly, but the adoption event clearly differs between N. fusca and Maculinea. Niphanda fusca lay eggs in small batches near the aphid colonies attended by their host ant, Camponotus japonicus, and the caterpillars initially feed on the aphid honeydew (Yamaguchi 1989). Caterpillars of N. fusca maintain contact with the host ant workers soon after their hatching. Nagayama (1950) suggested that the host workers adopted specifically only third-instar caterpillars, but Hirukawa (1978) observed that they also adopted second-instar caterpillars. We accordingly wished to test whether the adoption behavior of C. japonicus worker ants depends on the instars of N. fusca caterpillars. The cuticular chemical comparison was also conducted to investigate the possibility of chemical mimicry of cuticular hydrocarbons.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Collection and rearing of focal species

We collected approximately 100 eggs of N. fusca with aphids (species not identified) feeding on Japanese pampas grass Miscanthus sinensis and the tending C. japonicus colonies with workers, broods, males and gynes but no mature queen, at Fujinomiya in Shizuoka prefecture, Japan, in 2005 and 2011. In the laboratory, newly hatched caterpillars were reared with the aphids. Colonies of C. japonicus were reared at room temperature and under natural light conditions in a plastic box (350 mm × 250 mm × 60 mm) serving as a foraging arena, in which two nest boxes (110 mm × 75 mm × 30 mm) had been placed. Each nest box was covered with a glass plate (120 mm × 90 mm × 3 mm) to allow observation. Mealworms and 10% sucrose aqueous solution were provided as food in the arena box two times per week.

Adoption experiments

One day before the experiments, the adoption arena (45 mm × 80 mm) was connected with a Tygon tube to the nest box of C. japonicus. When the caterpillars grew to the second or third instars, 5 of them (<24 h after molting) were introduced into the adoption arena at the end of the foraging arena furthest from the nest entrance. Caterpillars were aggregated, as has been often observed in the field (M K Hojo, pers. obs., 2005). In the adoption arena, approximately 0–5 workers were foraging and 1 or 2 workers were inspecting the aggregated caterpillars. We measured antennation time (time being inspected by at least one worker) immediately after the first contact of the workers with one of the caterpillars for 5 min. We also observed the behavioral interaction between the workers and the caterpillars. This observation series was replicated 10, 30, 60 and 120 min after the first measurement. We also measured the adoption time from the time at which the caterpillar was placed in the foraging arena until it was brought to the connecting tube into the nest chamber by a worker ant. The experiments were replicated 6 times with 6 different colonies. Thus, 30 each of the second- and third-instar caterpillars were used. In each experimental set, when the caterpillars were adopted, 2 of the 5 third-instar caterpillars were recovered using forceps as soon as possible and subjected to chemical analysis. Two second-instar caterpillars were also recovered after 48 h and used for chemical analysis. The rest of the caterpillars were later adopted by the ants and used for other studies. All experiments were conducted during the daytime. In each colony, the experiments for second- and third-instar caterpillars were conducted randomly. Antennation time was analyzed using a linear mixed model. A model was constructed using antennation time as the response variable with an identity link function, the experimental set as a random effect, and the instars, time after first contact, and their interaction as fixed effects. The influences of variables were tested using a likelihood ratio test.

Chemical analyses

Second- and third-instar caterpillars were each immersed in 20 μL of n-hexane for 5 min. A host worker was also extracted in the same manner to confirm shared hydrocarbons between the host and the caterpillars. The extract was applied to a silica gel column, and hydrocarbons were eluted in n-hexane. The sample was concentrated, and a 0.5-caterpillar equivalent was applied to a gas chromatograph (GC-14A; Shimadzu, Kyoto, Japan) equipped with a DB-1 nonpolar capillary column (30 m × 0.25 mm × 0.25 μm; J & W Scientific, Folsom, CA, USA) and a flame ionization detector. Data were collected and calculated using CR-6A Chromatopac data processor (Shimadzu). Representative samples were used for GC-MS analyses. GC-MS analyses were performed using an HP5890-II GC (Hewlett Packard, Palo Alto, CA, USA) interfaced to a JEOL SX102A double-focusing magnetic sector mass spectrometer (JEOL, Tokyo, Japan) in the EI mode with 70 eV. Bray–Curtis similarities, computed with fourth-root transformed relative abundances of components, were used to compare hydrocarbon profiles. The similarity matrix of pairwise comparisons among all samples was represented in a nonmetric multidimensional scaling (NMDS) plot, and differences between groups were tested using an analysis of similarity (anosim). Details of the chemical analysis are described in Hojo et al. (2009). All statistical analyses were performed using R v2.12.1 software (R Development Core Team 2011).

Observation of caterpillar morphology

Before the behavioral experiment in 2005, each caterpillar was anesthetized on ice. Under a stereomicroscope fitted with an ocular micrometer, the presence of exocrine glands (ant organs) was assessed and the maximum head capsule width was measured.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Morphological observations revealed that the third-instar caterpillars had larger head capsules than the second-instar caterpillars (third instar: 0.36 ± 0.031, second instar: 0.32 ± 0.024, mean ± SD (mm), n = 15, Student's t-test, t = −4.73, P < 0.001). Two morphologically distinct structures were found on the dorsal surface of the third-instar caterpillars. Abdominal segment 8 had tentacle organs (TOs) consisting of paired eversible tubercles (Fig. 1B,D; black arrowheads). The third-instar caterpillars also had a dorsal nectary organ (DNO) with a distinct transverse opening on their abdominal segment 7 (Fig. 1B,D; white arrowhead). However, TOs and DNO were absent in the dorsal surface of the second-instar caterpillars (Fig. 1A,C). We could not confirm the presence and distribution of other ant organs such as perforated cupola organs and dendritic setae at this magnification.

figure

Figure 1. Morphological differences between (A,B) second- and (C,D) third-instars of Niphanda fusca. Black and white arrowheads indicate the tentacle organs and the dorsal nectary organ, respectively. Scale bars, 0.5 mm.

Download figure to PowerPoint

Of the 30 third-instar caterpillars, 27 (90%) were successfully adopted by C. japonicus workers, but only two second-instar caterpillars (0.7%) were adopted during 48 h of observation (Fisher's exact test using the sum of 6 colonies, P < 0.0001). The mean adoption time was 130.44 ± 63.63 min for the third-instar caterpillars (n = 27, mean ± SD) and 127 and 189 min for the second-instar caterpillars. In contrast to the adoption behavior, the antennation time of worker ants toward caterpillars did not differ significantly between the second and third instars (instar, χ2 = 0.023, d.f. = 1, P = 0.878; instar × time, χ2 = 0.150, d.f. = 1, P = 0.697; Fig. 2). When they antennated the third-instar caterpillars, the workers occasionally consumed the nectar secreted from DNO and rushed out in response to eversion of TOs, but they did not show such behaviors toward the second-instar caterpillars.

figure

Figure 2. Mean antennation times of Camponotus japonicus workers toward caterpillars of Niphanda fusca (n = 6 in each instar). Antennation times toward third-instar caterpillars did not differ significantly from those toward second-instar caterpillars. Error bars, standard errors. Black and white circles indicate second- and third-instar caterpillars, respectively.

Download figure to PowerPoint

Chemical analyses of the caterpillars detected 40 peaks, 31 of which were tentatively identified. Niphanda fusca caterpillars showed both C. japonicus-like (9 peaks) and N. fusca-specific (31 peaks) hydrocarbons on their cuticles (Table 1; Fig. 3). The NMDS plot tended to overlap between instars, and we could not statistically separate the second- and third-instar caterpillars of N. fusca (anosim, R = 0.1372, P = 0.055; Fig. 4).

figure

Figure 3. Gas chromatograms (flame ionization detector (FID)) of cuticular compounds from (A) second- and (B) third-instar Niphanda fusca and (C) host Camponotus japonicus workers. Numbers on the representative peaks correspond to the compounds in Table 1.

Download figure to PowerPoint

figure

Figure 4. Nonmetric multidimensional scaling (NMDS) plot based on the relative proportions of 40 cuticular hydrocarbons from second- and third-instar Niphanda fusca. Black and white circles indicate second- and third-instar caterpillars, respectively.

Download figure to PowerPoint

Table 1. Relative amounts of cuticular hydrocarbons of Niphanda fusca
Peak no.CompoundRetentin indexRelative amount (mean ± SD)
Second-instarThird-instarP valueHost ant
  1. Mann–Whitney U-test; n = 12 in each instar. Bold peak numbers indicate hydrocarbons common to Camponotus japonicus workers (n = 15). –, not detected.

 1n-C23:122.770.17 ± 0.08
2n-C2323.000.65 ± 0.590.47 ± 0.230.9320.33 ± 0.26
 3n-C25:124.740.39 ± 0.09
4n-C2525.003.45 ± 1.344.26 ± 2.630.7120.71 ± 0.28
 5n-C26:125.721.61 ± 0.88
6n-C2626.000.35 ± 0.110.55 ± 0.270.0682.11 ± 0.51
 75,7,12-trimeC2526.220.36 ± 0.30
 8n-C27:1(9)26.7524.86 ± 2.47
9n-C27:1(7)26.820.42 ± 0.360.61 ± 0.350.1008.65 ± 1.03
10n-C2727.001.29 ± 0.591.30 ± 0.610.7125.77 ± 0.94
1113-meC2727.371.69 ± 0.75
125-meC2727.560.67 ± 0.36
137,15-dimeC2727.761.82 ± 2.011.37 ± 1.330.7126.00 ± 2.15
14n-C2828.000.89 ± 0.231.01 ± 0.240.3181.14 ± 0.23
155,7,12-trimeC2728.220.96 ± 0.82
16unknown28.310.78 ± 0.522.05 ± 2.470.127
17unknown28.630.82 ± 0.330.82 ± 0.260.977
18n-C29:127.770.87 ± 0.360.82 ± 0.330.75542.31 ± 7.95
19n-C2929.000.76 ± 0.240.83 ± 0.390.9772.19 ± 0.47
20unknown29.631.09 ± 0.211.08 ± 0.460.798
21dimeC2929.740.34 ± 0.580.85 ± 0.740.068
22n-C3030.002.00 ± 0.803.08 ± 2.070.197
23meC3030.180.50 ± 0.150.52 ± 0.410.589
242-meC3030.6237.28 ± 4.8634.19 ± 8.830.477
25dimeC3030.810.86 ± 0.331.12 ± 0.980.977
26n-C3131.005.59 ± 1.994.58 ± 1.530.143
27dimeC3131.620.91 ± 0.280.89 ± 0.241.000
28n-C3232.004.79 ± 3.225.88 ± 6.450.712
29meC3232.200.70 ± 0.160.78 ± 0.340.551
30unknown32.290.30 ± 0.130.37 ± 0.110.088
314-meC3232.517.40 ± 2.095.68 ± 1.830.068
32unknown32.821.20 ± 0.511.36 ± 0.840.798
33dimeC3232.883.57 ± 1.292.82 ± 0.650.100
34unknown33.180.46 ± 0.210.54 ± 0.280.589
3513-meC3333.281.10 ± 0.711.07 ± 0.560.977
36dimeC3333.540.79 ± 0.260.93 ± 0.560.377
37n-C3434.004.86 ± 1.815.55 ± 1.980.171
38unknown34.270.66 ± 0.290.50 ± 0.160.241
39dimeC3434.600.34 ± 0.160.27 ± 0.160.347
4013-meC3535.271.86 ± 0.981.55 ± 0.910.513
4113,17-dimeC3535.551.01 ± 0.701.05 ± 0.480.477
42n-C3636.005.15 ± 1.465.28 ± 1.650.842
43unknown36.230.51 ± 0.210.47 ± 0.200.629
44unknown36.890.52 ± 0.300.63 ± 0.250.197
4515-meC3737.191.12 ± 0.521.29 ± 0.600.442
4610,13-dimeC3737.481.11 ± 0.551.22 ± 0.570.551
4713,17-meC3939.150.73 ± 0.310.96 ± 0.410.160
4813,16-dimeC3939.430.96 ± 0.461.19 ± 0.500.218

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We confirmed that the adoption of the caterpillars by host ants depended on the caterpillar instar. Host ants adopted most third-instar caterpillars but not adopt most second-instar caterpillars. Despite clear differences in ant adoption based on the caterpillar instars, chemical analysis revealed the similarity of the cuticular hydrocarbon profiles among the caterpillars. This finding suggests that the mimetic cuticular hydrocarbons alone are not sufficient as the adoption signal in N. fusca. Indeed, our behavioral assays indicated that the adoption time of N. fusca caterpillars is much longer than that of two other parasitic butterflies, M. rebeli and M. alcon (approximately 10–30 min in the laboratory: Akino et al. 1999; Als et al. 2001), although conditions were not exactly the same. This discrepancy in adoption times may be due to differences in the chemical strategies employed by the parasitic caterpillars. The blue butterflies M. rebeli and M. arion are cuckoo feeders in host Myrmica ant brood chambers, and the host workers adopt the parasitic caterpillars a few minutes after first contact (Akino et al. 1999; Als et al. 2001). The parasitic M. rebeli caterpillars enable this adoption by mimicking the hydrocarbon of host ant larvae, and close matching of cuticular chemistry is crucial for inducing rapid adoption behavior (Akino et al. 1999; Nash et al. 2008). In contrast to Maculinea, N. fusca adults oviposit near aphid colonies with clutches, and ant–aphid mutualism is exploited so that the hatched caterpillars are tended by the host ants for 1–2 weeks. In such a situation, N. fusca caterpillars can use different manipulative signals rather than the mimicry of hydrocarbons. Once adopted, however, N. fusca caterpillars mimic the colony- and caste-specific hydrocarbon profiles of the host ant (Hojo et al. 2009). In our experiments, adopted second-instar caterpillars were not killed but were tended by the workers in the ant nest. Thus, even if caterpillars were adopted as second instars, they would be integrated into the host colony through the chemical mimicry.

Lycaenid caterpillars possess various specialized ant organs used in communication with ants (Pierce et al. 2002). The duration of antennation behavior by C. japonicus workers did not differ between the second and third instars of N. fusca (Fig. 1). However, the behavioral repertories of the workers following antennation differed slightly, such that the workers occasionally exhibited rushing behavior toward everted TOs and consumed DNO secretions when they antennated the third-instar caterpillars but did not show such behavior toward second-instars. DNO secretes nutritious droplets that are specifically attuned to the taste of C. japonicus (Hojo et al. 2008), and TOs are presumed to emit volatile compounds that induce alarm behavior in the ants (Pierce et al. 2002). Thus, chemical signals emitted from TOs and DNO could be involved in adoption events in N. fusca.

Recent studies have shown that ants can learn to associate cuticular chemicals with food rewards (Bos et al. 2010, 2012) so that the same cuticular cue might carry a different meaning for the ants depending on the context. Although similar chemical (hydrocarbon) cues were presented on the second- and third-instar caterpillars, the ants received these chemicals in a different context (i.e. presence or absence of food rewards and TOs). The importance of feeding on DNO secretion and rushing behavior during the adoption event was also noted in M. arion (Thomas 2002) and the African lycaenid butterfly Aloeides dentatis (Henning 1983) but experimental evidence for the existence of an adoption signal from DNO and TOs is still lacking in all of these species. In addition to chemical signals, recent studies have indicated that acoustic signaling is used in parasitic lycaenid–ant interactions (Barbero et al. 2009a,b). It is probable that a combination of multimodal signals from ant organs and cuticular hydrocarbons is involved in adoption events.

In this study, most of the second-instar caterpillars were not adopted by the host workers, but in field observations of Hirukawa (1978), most of caterpillars were adopted at the second larval instar in the wild. One of the important differences between the wild and our experimental conditions is the presence of aphids that are also tended by the workers. Because the aphid's honeydew and alarm pheromones affect the behavior of the tending ants (Verheggen et al. 2012), it is possible that the ant-tended aphids play a role in the adoption of N. fusca caterpillars at certain conditions. Further behavioral assays focusing on multimodal signals and multi-specific interaction will shed light on the evolution of lycaenid–ant communication.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We thank Ayako Wada-Katsumata and Mamiko Ozaki for many helpful suggestions regarding this study. This work was supported in part by a Research Fellowship of the Japan Society for the Promotion of Science for young scientists (No.223146) to M.K.H.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • Akino T (2008) Chemical strategies to deal with ants: a review of mimicry, camouflage, propaganda, and phytomimesis by ants (Hymenoptera: Formicidae) and other arthropods. Myrmecological News 11, 173181.
  • Akino T, Knapp JJ, Thomas JA, Elmes GW (1999) Chemical mimicry and host specificity in the butterfly Maculinea rebeli, a social parasite of Myrmica ant colonies. Proceedings of the Royal Society B 266, 14191426.
  • Als TD, Nash DR, Boomsma JJ (2001) Adoption of parasitic Maculinea alcon caterpillars (Lepidoptera: Lycaenidae) by three Myrmica ant species. Animal Behaviour 62, 99106.
  • Barbero F, Bonelli S, Thomas JA, Balletto E, Schönrogge K (2009a) Acoustical mimicry in a predatory social parasite of ants. Journal of Experimental Biology 212, 40844090.
  • Barbero F, Thomas JA, Bonelli S, Balletto E, Schönrogge K (2009b) Queen ants make distinctive sounds that are mimicked by a butterfly social parasite. Science 323, 782785.
  • Bos N, Dreier S, Jørgensen CG, Nielsen J, Guerrieri FJ, d'Ettorre P (2012) Learning and perceptual similarity among cuticular hydrocarbons in ants. Animal Behaviour 58, 138146.
  • Bos N, Guerrieri FJ, d'Ettorre P (2010) Significance of chemical recognition cues is context dependent in ants. Animal Behaviour 80, 839844.
  • Henning SF (1983) Chemical communication between lycaenid larvae (Lepidoptera: Lycaenidae) and ants (Hymenoptera: Formicidae). Journal of the Entomological Society of Southern Africa 46, 341366.
  • Hirukawa N (1978) [Life history of Niphanda fusca (Bremer et Grey) (Lepidoptera, Lycaenidae) in Kiso-dani, Nagano Prefecture, 1]. Insects and Nature 13, 118. (In Japanese.)
  • Hojo MK, Wada-Katsumata A, Akino T, Yamaguchi S, Ozaki M, Yamaoka R (2009) Chemical disguise as particular caste of host ants in the ant inquiline parasite Niphanda fusca (Lepidoptera: Lycaenidae). Proceedings of the Royal Society B 276, 551558.
  • Hojo MK, Wada-Katsumata A, Ozaki M, Yamaguchi S, Yamaoka R (2008) Gustatory synergism in ants mediates a species-specific symbiosis with lycaenid butterflies. Journal of Comparative Physiology A 194, 10431052.
  • Martin SJ, Drijfhout FP (2009) A review of ant cuticular hydrocarbons. Journal of Chemical Ecology 35, 11511161.
  • Martin SJ, Helanterä H, Drijfhout FP (2008) Colony-specific hydrocarbons identify nest mates in two species of Formica ants. Journal of Chemical Ecology 34, 10721080.
  • Nagayama H (1950) [Life history of Niphanda fusca Bremer et Grey]. The Insect Ecology 3, 918. (In Japanese.)
  • Nash DR, Als TD, Maile R, Jones GR, Boomsma JJ (2008) A mosaic of chemical coevolution in a large blue butterfly. Science 319, 8890.
  • Pierce NE, Braby MF, Heath A et al. (2002) The ecology and evolution of ant association in the Lycaenidae (Lepidoptera). Annual Review of Entomology 47, 733771.
  • R Development Core Team (2011) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. [Cited 26 September 2012.] Available from URL: http://www.R-project.org.
  • Schlick-Steiner BC, Steiner FM, Höttinger H et al. (2004) A butterfly's chemical key to various ant forts: intersection-odour or aggregate-odour multi-host mimicry? Die Naturwissenschaften 91, 209214.
  • Thomas JA (2002) Larval niche selection and evening exposure enhance adoption of a predacious social parasite, Maculinea arion (large blue butterfly), by Myrmica ants. Oecologia 132, 531537.
  • Thomas JA, Schönrogge K, Elmes GW (2005) Specializations and host associations of social parasites of ants. In: Fellowes MDE , Holloway GJ , Rolff J (eds) Insect Evolutionary Ecology, pp 479518. CABI Publishing, Wallingford.
  • Verheggen FJ, Diez L, Sablon L et al. (2012) Aphid alarm pheromone as a cue for ants to locate aphid partners. PLoS ONE 7, e41841.
  • Yamaguchi S (1989) [The Life Histories of Five Myrmecophilous Lycaenid Butterflies of Japan]. Kodansha, Tokyo. (In Japanese.)