Perception of oviposition-deterring larval tracks in aphidophagous coccinellids Cycloneda limbifer and Ceratomegilla undecimnotata ( Coleoptera : Coccinellidae )

The ability of the aphidophagous coccinellids Cycloneda limbifer Casey and Ceratomegilla undecimnotata (Schneider) to discriminate between simultaneously provided clean paper strips and paper strips with oviposition-deterring larval tracks was studied after the ablation of different sense organs. Females oviposited similar numbers of eggs on paper strips with conspecific tracks and on clean paper strips only when deprived of both maxillary palpi. C. undecimnotata without maxillary palpi also did not differentiate between clean paper strips and paper strips with tracks of the coccinellid Leis dimidiata (F.). If both antennae and one maxillary palpus were simultaneously ablated, females of both species laid significantly more eggs on clean than contaminated paper strips. The results of this study indicate that females use contact chemoreceptors on maxillary palpi exclusively to detect oviposition deter­ ring tracks of conspecific larvae. Intact females of C. limbifer laid significantly larger batches of eggs on paper strips with conspecific larval tracks, than on clean paper strips in blank test. In contrast, intact females of C. undecimnotata laid significantly smaller batches on paper strips with conspecific tracks than on clean paper strips in blank test. This is the first evidence of an opposite effect of conspecific oviposition deterring larval tracks on egg clustering in aphidophagous coccinellids.

The location of various receptors on sense organs has been studied in many species of aphidophagous coccinel-lids.In Coccinella septempunctata L., chemoreceptors were described on the labial and maxillary palpi (Yan et al., 1982(Yan et al., , 1987)).In Ceratomegilla undecimnotata (Schneider), gustatory and mechanosensitive sensilla organs were reported along the entire length of the anten nae, while olfactory sensillae were confined to two ter minal segments of the antennae (Jourdan et al., 1995).On the maxillary palpi of aphidophagous coccinellids, olfac tory chemoreceptors were more numerous than gustatory ones.The total number of both was always higher in polyphagous than in oligophagous or monophagous spe cies, the numbers being even lower in coccidophagous coccinellids, which feed on immobile prey (Barbier et al., 1989(Barbier et al., , 1996)).In addition, some mechanoreceptors were also described on the palpi.Obata (1986) suggested that the aphidophagous coccinellid Harmonia axyridis (Pallas) used olfactory stimuli to locate sites with aphids for the first time.Adult C. septempunctata respond to aphid alarm pheromone via olfactory cells (Al Abassi et al., 2000).This species was attracted to aphid odours in olfactometer trials and even differentiated between vola tiles of different aphid species (Sengonca & Liu, 1994).Aphid released volatiles also attracted adults of Hippodamia convergens Guérin-Méneville (Acar et al., 2001), but adults without antennae or antennal tips did not respond to volatile cues (Hamilton et al., 1999).
Laboratory experiments showed that a volatile oviposition-deterring substance, which evaporated from larval tracks of the aphidophagous chrysopid Chrysopa oculata Say within a few days, can contaminate clean substrates.In contrast, sites with tracks of larvae pre served their deterrent effects on females for much longer time (Růžička, 1997a).Intraspecific oviposition-deterring effects of tracks of first instars of the coccinellid C. septempunctata disappeared after one day (Růžička, 1997b), whereas tracks of some other coccinellid species effec tively deterred females from oviposition for many days (Růžička, 2002).However, unpublished results of labora tory experiments indicated the absence of a volatile spacing cue in the tracks of their larvae.Females of some coccinellids also avoided oviposition on sites with tracks ofheterospecific coccinellid larvae (Růžička, 2001).
Experiments in this study were designed to identify organs with chemoreceptors for the perception of oviposition-deterring semiochemicals in females of C. limbifer and C. undecimnotata.Both species avoided ovi positing on sites with larval tracks of some heterospecific coccinellids, but the effect of tracks of L. dimidiata larvae on C. undecimnotata was the most intense.The ability to differentiate between simultaneously provided clean and ODS contaminated sites was studied in laboratory experi ments with females deprived of different sense organs.

MATERIAL AND METHODS
Experiments were done on Cycloneda limbifer Casey, Ceratomegilla undecimnotata (Schneider) and Leis dimidiata (F.), from laboratory cultures.The oviposition behaviour of normal females in choice tests with clean and contaminated sites was compared with the oviposition behaviour of females after the ablation of different sense organs.The ability of females to avoid sites with conspecific tracks was studied in experiments with C. limbifer and C. undecimnotata.In addition, the ability of C. undecimnotata females to discriminate between clean sites and sites with tracks of L. dimidiata larvae after these ablations was investigated.Transversely folded paper strips, one clean and one previously exposed to larvae, were provided for ovipo siting simultaneously.Each strip was 20 cm long and 4 cm wide, the length of folded paper strip was 13 cm.Ten females without males were present in the test simultaneously.Each test was terminated after 20 hours and had ten replicates.
Ablation experiments.Two tests were made with intact females, a control choice test (0) with one clean and one con taminated substrate, and a blank oviposition test (BT) with two clean substrates.Oviposition in these tests was compared with oviposition of females deprived of different organs, which may detect any oviposition deterring semiochemicals.Prior to the ablation, each coccinellid female was carefully fixed with narrow strips of scotch tape to small piece of paper.All seg ments of amputated organ(s) were then removed under bin ocular glass by means of fine surgical instruments.Either one or both sense organs of one kind, or different combinations of these organs, were eliminated: one antenna (1a), both antennae (2a), one leg of the second pair (1l), one antenna and one maxil lary palpus (1a1p), two antennae and one maxillary palpus (2a1p), two maxillary palpi (2p), one antenna and two maxil lary palpi (1a2p), two antennae and two maxillary palpi (2a2p).
Contam ination of substrates by larvae.Two folded paper strips were exposed to 40 unfed first-instar larvae of one species for four hours on the bottom of a glass Petri dish, 18 cm in diameter.The inner wall of the dish was painted with Fluon (Polytetrafluoroethylene (PTFE) Dispersion; Imperial Chemical Industries Limited, Plastics Division, Welwyn Garden City, Herts, UK), to prevent dispersal of larvae from the bottom.Con taminated paper strips were offered for egg laying immediately after their exposure to larvae.
Oviposition choice test.The design of choice tests was adopted from previous studies on aphidophagous coccinellids (Růžička, 1997b;2001).The cylindrical experimental cage, with substrates, was 10 cm high, 18 cm in diameter.The side of the cage was made from firm netting, the bottom and the top was made from glass.Two folded paper strips, one clean and one previously exposed to larvae, were provided for oviposition to 10 females for 20 hours.Symmetrically spaced strips were on the bottom, 50 mm apart.Drinking water and a surplus of suit able aphid prey, Aphis fabae Scopoli for C. limbifer and Acyrthosiphum pisum Harris for C. undecimnotata, were avail able to females during the tests.A. fabae was offered on shoots of the host plants, Viciafaba, A. pisum without the host plant.Each test had ten replicates.Experiments were carried out at 24 ± 2°C with a 18L : 6D photoperiod, using light from fluorescent tubes.
Statistical analysis.Differences between numbers of eggs on two paper strips in each test were compared using the Wilcoxon paired sample test.The same test was used to compare differ ences between numbers of eggs on and outside paper strips.Size differences between clusters of eggs on contaminated paper strips in choice tests and on clean paper strips in blank test were compared using the Student's t-test.

Effects of ablations on female discrimination between clean and contaminated paper strips
Intact females of C. limbifer and C. undecimnotata laid similar numbers of eggs on two clean paper strips in blank tests with conspecific larvae (P = 1 and P = 0.4316), and C. undecimnotata in blank test with larvae of L. dimidiata (P = 0.5566).Intact females of both spe cies effectively avoided oviposition on sites with larval tracks in the choice test with conspecific larvae (P < 0.01 and P < 0.01), and C. undecimnotata in the choice test with tracks ofheterospecific larvae (P < 0.01) (Fig. 1).
Females of C. limbifer also laid significantly lower numbers of eggs on paper strips with tracks of conspe cific larvae than on clean paper strips after the ablation of one antenna (P < 0.05), both antennae (P < 0.05), one leg of the second pair (P < 0.01), one palpus maxillaris and one antenna (P < 0.01), one palpus maxillaris and both antennae (P < 0.01), or one palpus maxillaris (P < 0.05).Females without both maxillary palpi laid similar num bers of eggs on clean and contaminated paper strips (P > 0.9999), and also when one and both antennae were simultaneously ablated with maxillary palpi (P = 0.1484 and P = 0.8438) ( Fig. 1).

Effects of ablations on oviposition outside paper strips
Coccinellid females avoided oviposition outside paper strips in the control and blank tests with conspecific and heterospecific larvae (P < 0.01 for each test control and blank test) (Fig. 2).Females of C. limbifer without maxil lary palpi, or without maxillary palpi and one antenna, or maxillary palpi and both antennae, laid more eggs outside paper strips than on paper strips (P = 0.01 for each test).C. undecimnotata laid more eggs outside paper strips than on strips in tests with conspecific and heterospecific larval tracks when maxillary palpi (P < 0.01) [P < 0.01], or maxillary palpi and one antenna were ablated (P < 0.01) [P < 0.05], but females without maxillary palpi and both antennae laid their eggs almost exclusively on paper strips (P < 0.01) [P < 0.05] (Fig. 2).

Effects of ablations on egg clustering
C. limbifer laid egg batches of similar size on the two clean paper strips in the blank test (P = 0.7994).In the control test, intact females of C. limbifer also laid batches of similar size on the clean and contaminated paper strips (P = 0.8229); however, the average cluster size on these paper strips was significantly larger than the average size of clusters on the paper strips in the blank test (P < 0.0001).Batches of eggs oviposited on contaminated paper strips were also larger than those laid on clean paper strips in the blank test when females were deprived of one antenna (P < 0.0001), both antennae (P < 0.0001), one palpus maxillaris (P < 0.0001), and one palpus maxillaris and both antennae simultaneously (P = 0.05).
C. undecimnotata also laid egg batches of similar size on the two clean paper strips in the blank tests in the experiments with conspecific and heterospecific larvae (P = 0.5156) [P = 0.2241].However, C. undecimnotata females laid significantly smaller batches on paper strips with conspecific tracks in the control test than on clean paper strips in the blank test (P < 0.01).In the control test with larval tracks of L. dimidiata, the size of batches on the paper strips with tracks did not differ from the size of batches on clean paper strips in the blank test [P = 0.2041] (Fig. 3).

DISCUSSION
Coccinellid females are voracious and relatively mobile aphid predators.When searching for the prey and oviposition sites, they must frequently come across the oviposition deterring larval tracks of different competitors for the same prey species.Females are likely to detect these tracks through direct contact.This study showed that females of both coccinellid species, C. limbifer and C. Choice tests with intact females (0), females without: one antenna (1a), both antennae (2a), one leg of the second pair (1l), one antenna and one maxillary palpus (1a1p), both antennae and one maxillary palpus (2a1p), one maxillary palpus (1p), both maxillary palpi (2p), one antenna and both maxillary palpi (1a2p) both antennae and both maxillary palpi (2a2p), and blank test with intact females (BT).Wilcoxon paired sample test, ** = P < 0.01; * = P < 0.05; ns = not significantly different (P > 0.05).
palpi.Without both maxillary palpi, females of each spe cies were unable to discriminate clean paper strips from paper strips with oviposition-deterring larval tracks of conspecific larvae.After the ablation of maxillary palpi, C. undecimnotata females were unable to discriminate also between clean sites and sites with larval tracks of L. dimidiata.Females with one maxillary palpus effectively avoided oviposition on sites previously exposed to conspecific or heterospecific larvae.Both species success fully discriminated between clean and contaminated sites also when one maxillary palpus was ablated simultane ously with one or both antennae.This also indicated, that ablation injuries had generally small impact on female ability to avoid oviposition on the less suitable of two oviposition sites.Similarly, ablation of one leg of the second pair had little impact on the ability of females of both coccinellids to discriminate between sites with and without larval tracks.Obata (1986Obata ( , 1997) ) reported that adults of Harmonia axyridis (Pallas) used olfactory cues to find aphids.In C. septempunctata, Sengonca & Liu (1994) confirmed that coccinellid adults can be attracted by aphid odours.Zhu et al. (1999) recorded strong electroantennogram responses from antennae to a hexane extract of the vola tile semiochemicals associated with the prey species and its host plant in Coleomegilla maculata lengi (Timberlake).Ablation of the antennae in adults showed that H. convergens perceives volatile semiochemicals associated with their prey through this organ (Hamilton et al., 1999).The contact perception of ODS by maxillary palpi alone in C. limbifer and C. undecimnotata could also explain why A. bipunctata did not inhibit their ovipo sition, when females were exposed to the odour of conspecific larvae, but had not access to the surface with larval tracks (Hemptinne et al., 1992).
Intact females of both species laid eggs mostly on the lower side of the clean paper strips provided for egg lay ing.Females with one maxillary palpus also strongly pre ferred oviposition on the paper substrate.C. limbifer laid almost all eggs outside the paper strips after the ablation of both maxillary palpi and when both palpi with one or both antennae were simultaneously removed.C. undecim notata without maxillary palpi, or maxillary palpi and one antenna, preferred oviposition outside strips also, but laid a larger portion of eggs on paper strips than females of C. limbifer.In contrast to C. limbifer, C. undecimnotata laid most eggs on the paper strips than outside when both palpi and both antennae were ablated.Current knowledge does not explain why females changed their usual prefer ence for paper strips, after the ablation of both maxillary palpi in three choice tests with C. limbifer, and in two out of three tests with C. undecimnotata.This change of ovi position behaviour, and the difference between species, might be associated with the location of mechanoreceptors on the ablated sense organs.
Comparison of intra-and interspecific oviposition deterring effects of larval tracks in aphidophagous cocci nellids (Růžička, 2001) indicated the occurrence of dif ferent semiochemicals, or different ratios of active compounds, in larval tracks of different species.Although females of C. limbifer effectively avoided oviposition on sites with conspecific larval tracks and C. undecimnotata larval tracks, fresh tracks of conspecific larvae had a strong attractant effect on their searching behaviour, while heterospecific tracks strongly deterred females from searching (Růžička & Zemek, in press).Because only intact females were used in this study, sense organs which perceived cues modifying searching behaviour of females were not identified.
Many predatory arthropods lay eggs in clusters.The clutch size in coccinellids was positively correlated with the number of ovarioles (Stewart et al., 1991).Within a coccinellid species, the size of cluster seemed to be directly related to aphid density (Agarwala & Bardhanroy, 1999).Egg clustering in mites can be an adaptation to counter attack by prey (Faraji et al., 2002), in coccinellids to interspecific predation (Agarwala & Dixon, 1993;Agarwala & Yasuda, 2001).This study revealed a strong opposite effect of conspecific larval tracks on egg cluster size in studied species.Females of C. limbifer with all sense organs also laid significantly larger batches on clean and contaminated paper strips in the control test than on clean paper strips in the blank test.Females with both maxillary palpi also laid larger batches in the choice tests than in the blank test when deprived of one or both antennae.Similar behaviour was observed in females with one maxillary palpus or females with one maxillary palpus and without antennae.The increase of clutch size on sites with conspecific tracks in C. limbifer might be related to attractant effects of conspecific tracks on female searching behaviour described in the previous paragraph (Růžička & Zemek, in press).In contrast, intact females of C. undecimnotata laid significantly smaller egg batches on the paper strips with conspecific tracks than on clean paper strips in the blank test.This differ ence between the oviposition behaviour of C. limbifer and C. undecimnotata females on sites with ovipositiondeterring tracks of conspecific larvae and on clean sites indicates opposite kinds of adaptation of these coccinellid species to the presence of conspecific competitors.This is an interesting phenomenon in oviposition strategies of closely related predators.Presumably, both tactics can successfully enhance progeny survival.
This study showed that aphidophagous coccinellids C. limbifer and C. undecimnotata used exclusively contact chemoreceptors on maxillary palpi to detect ovipositiondeterring tracks of conspecific and heterospecific larvae.It also disclosed that sites with deterrent conspecific tracks effectively stimulate females of C. limbifer to lay larger clusters of eggs, while C. undecimnotata females tend to lay smaller egg clusters on sites with conspecific tracks than on sites without tracks.The opposite response of females to the presence of larval competitors is a new element among adaptations in oviposition strategies of aphidophagous predators.