Chem. Senses 26: 983-992,
2001
© Oxford University Press 2001
Taste-enhancing Effects of Glycine on the Sweetness of Glucose
a Gustatory Aspect of Symbiosis between the Ant, Camponotus japonicus, and the Larvae of the Lycaenid Butterfly, Niphanda fusca
Department of Applied Biology, Faculty of Textile Science, Kyoto Institute of Technology, Matsugasaki, Sakyo-ku, Kyoto 606-8585, Japan 1 Department of Biology, Faculty of Science, Nara Women's University, Nara 630-8506, Japan 2 205 Lakeside, 3-2-8 Ishikami-chyo, Nerima-ku, Tokyo 177-0041, Japan
Correspondence to be sent to: Mamiko Ozaki, Department of Applied Biology, Faculty of Textile Science, Kyoto Institute of Technology, Kyoto 606-8585, Japan. e-mail: mamiko{at}ipc.kit.ac.jp
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The lycaenid butterfly, Niphanda fusca, has a parasitic relationship with its host ant, Camponotus japonicus: the caterpillars may use chemical mimicry to enter the ant nest where they are fed mouth-to-mouth by the adult ants until pupation. Nevertheless, larvae offer their host ants a nutritious secretion that contains 160 mM glucose and 43 mM glycine. Using glucose and glycine mixture as artificial secretions, we investigated the gustatory effect of glucose and/or glycine on the ants. Glycine induced neither feeding behavior nor gustatory response in the ants if its concentration was <500 mM. In the presence of glycine at the concentration in the secretion, however, the ants improved their preference to glucose, and the sugar receptor cell exhibited electrophysiological enhancement of response to glucose in a glycine-concentration-dependent manner. By adding glycine to glucose in their secretions, therefore, the butterfly larvae can manipulate the gustatory sense of the ants. The alluring taste of `glycine-flavored glucose' could motivate the host ants to feed the larvae and thereby receive the secretions as a reward. The taste enhancement created by the combination of sugar and amino acid may play a role in the evolution of the parasitic relationships of these insects. The taste-enhancing effect appears to be analogous to taste enhancement by `umami' substances in humans.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The caterpillars of many species of lycaenid butterflies associate with ants and these interactions can range from parasitism to mutualism (Atsatt, 1981
;
Henning, 1983;
Pierce, 1987;
Thomas et al., 1989;
Hölldobler
and Wilson, 1990; Fiedler
et al., 1996). One of the representative species in
Japan, Niphanda fusca (Nagayama,
1950), remarkably and in contrast to most other ant-parasitic
lycaenids, feeds on aphid honeydew instead of plants when young. They start to
produce a sweet secretion from a specialized endocrine gland called the dorsal
nectary organ (DNO) at the third instar stage. The ants, Camponotus
japonicus, carry the larvae of this age to their nests and care for them
over the winter months. The ants have a strong interest in the larval DNO
secretions while feeding and grooming (see
Figure 1A). They often tap on
the back of the larva to stimulate secretion from the DNO (see
Figure 1B). Once an ant has
taken a drop of secretion, it holds it with its labial and maxillary palps
(arrows in Figure 2A) and
shares it with other nestmates. Thus, the DNO secretion is thought to be
crucial in retaining the relationship between the butterfly larvae and the
ants (Wardlaw et al.,
2000). However, these behavioral observations have not been
coupled with studies of the sensory physiology of the ants and how they
perceive the taste of the secretion.
|
|
The secretions of N. fusca have only once analysed because of
difficulty in capturing the larvae and in collecting the small amounts of
secretion. The analysis of Nomura et al.
(Nomura et al., 1992)
showed that the secretion contains 160 mM glucose (75% of the total sugar) and
43 mM glycine (76% of the total amino acid). It has been reported that the
larvae of other myrmecophilous butterflies, Jalmenus evagoras
(Pierce, 1989;
Pierce and Nash, 1999) and
Lysandra hispana (Maschwitz
et al., 1975) secrete serine and methionine instead of
glycine, respectively (see Discussion). As for the sugar component, a separate
study found that the secretion of N. fusca contained 600 mM glucose
(45% of the total sugar) (D. Chogyoji, personal communication). This
difference in concentration of glucose may be due to differences in the
nutritional states of the colonies of host ants feeding the larvae of N.
fusca.
Based on these data, we conducted behavioral and electro-physiological experiments on the taste of the DNO secretion of N. fusca, hypothesizing that the glucose—glycine combination in the DNO secretion could provide an especially alluring taste for workers of C. japonicus.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The ants
The ants, C. japonicus (Hymenoptera: Formicidae), were all collected from the same colony in the field. In our laboratory, they were kept in plastic boxes (350 x 225 x 55 mm) with only a supply of water. We subjected them to electrophysiological experiments 2 days after collection. Six days after collection, we used them for the behavior experiments, because the period of food deprivation increases the feeding sensitivity to be analysed. As the median of feeding threshold, K1/2, differed among groups of ants (see Figures 4A and 5A), data from a series of behavioral experiments were obtained in the same population.
|
|
Behavioral tests
We conducted two types of behavior tests: (i) choice-and-drink tests to examine preference between two kinds of test solutions and (ii) feeding response tests to determine the feeding threshold of a test solution.
The choice-and-drink test was carried out according to the method of
Tanimura et al. (Tanimura et
al., 1982), which was originally designed for fruit flies. In
order to determine the relative effects of glucose and glucose plus glycine,
we used three test plates, each of which had 60 wells. In each plate, the 60
wells were alternately filled with 10 µl of glucose solution and
glucose-plus-glycine solution. In plate 1, the glucose solution was stained
with a blue food color (brilliant blue FCF) and the glucose-plus-glycine
solution was colorless, so that the blue and the colorless wells made a
checkered pattern. In plate 2, colorless glucose and blue glucose-plus-glycine
solutions were used. In plate 3, both solutions were colorless. In preliminary
experiments, we established that the blue food color used in our study
affected neither electrophysiological nor feeding response to glucose. We
placed four ants onto each of these three plates and let them drink freely for
3 h in the dark. The ants were then frozen and their crops were isolated. The
crops were homogenized and extracted with 1 ml of 50 % ethanol. The absorbance
of the extracts was measured at 630 nm with a spectrophotometer. In order to
eliminate non-specific absorbance by the intrinsic crop extract, the
absorbance of the stomach extract in the ants of plate 3 was subtracted from
the absorbance of the crop extract of the ants of plates 1 and 2. Using the
compensated absorbancies of the extracts of plates 1 (Abs.1) and 2 (Abs.2),
the preference rates for glucose and glucose-plus-glycine solutions were
defined as Abs.1/(Abs.1 + Abs.2) and Abs.2/(Abs.1 + Abs.2), respectively. If
Abs.1/(Abs.1 + Abs.2) = Abs.2/(Abs.1 + Abs.2), the ants showed no difference
between glucose and glucose plus glycine. If Abs.1/(Abs.1 + Abs.2) >
Abs.2/(Abs.1 + Abs.2), the ants preferred glucose to glucose plus glycine. If
Abs.1/(Abs.1 + Abs.2) < Abs.2/(Abs.1 + Abs.2), the ants preferred glucose
plus glycine to glucose. Furthermore, we carried out similar experiments using
methionine and serine instead of glycine, because the major amino acid in the
DNO secretion is glycine in N. fusca, but serine has been reported to
be the main amino-acid component of the DNO secretion in Jalmenus
evagoras (Hunt et al.,
1982) and methionine has been found in trace amounts in the DNO
secretion of Lysandra hispana
(Maschwitz et al.,
1975).
For the feeding response tests, a glucose concentration series of the test solution was prepared (10 concentration steps using 1:2 dilutions with distilled water, starting from 2 M glucose) and three other glucose concentration series were prepared in the same way, but using 1:2 dilutions with 50, 100 or 200 mM glycine. Twenty ants were randomly chosen from a single colony and each was placed in a plastic pipette tip with the top cut to fit the head of the ant, so that the head of the ant stuck out of the opening. Thus restrained, the ants in the pipette tips were convenient to handle under the binocular microscope. A drop of test solution was placed under the microscope and carefully touched with the maxillary and the labial palps. Observing movements of the glossa, we quickly judged whether the ant was prepared to drink the test solution or not. We then counted the number of ants showing this feeding response with glossa extension. In drawing a concentration—feeding response curve based on this kind of experiment, the number of ants showing glossa extension was plotted against the logarithmically scaled concentration of glucose and curve fitting was done by computer.
Electrophysiological procedure
The isolated head of the ant was connected to an indifferent platinum
electrode. A glass capillary containing the stimulus solution also contained
the platinum wire of the recording electrode. Under the microscope, the long
chemosensillum on the second segment of the labial palps (arrow in
Figure 2B) was focused and the
capillary was slipped over the tip of the sensillum. Thus, the sensory
response was simultaneously recorded when the stimulus touched the sensillum
tip (Hodgson et al.,
1955). The stimulus solutions were prepared with various
concentrations of glucose and/or glycine dissolved in 100 mM NaCl. The 100 mM
NaCl solution alone did not induce significant responses (see
Figure 6B). The duration of
stimulation was 0.7-1 s and intervals of at least 3 min between stimuli were
taken to avoid any effects of adaptation. We used only workers with responsive
sensilla for these assays; a sensillum was considered to be responsive if it
generated >9 impulses/0.5 s to glucose in 100 mM NaCl.
|
Scanning electron microscope
The isolated heads of the ants were mildly sonicated in distilled water, dehydrated through the ethanol series of 70, 80, 85, 90, 95, 99 and 100% and finally treated with 100% acetone. They were then transferred into iso-amyl acetate. After they had been dried in the critical-point dryer, each specimen was put on a sample stage, coated with gold and observed with a scanning electron microscope (Hitachi S-2100A).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Effects of glycine on glucose-feeding behavior
In the choice-and-drink experiment
(Figure 3), we tested 400 mM
glucose versus 400 mM glucose plus 50 (A), 100 (B), or 200 mM glycine (C).
This concentration of glucose was the average value of measurements by Nomura
et al. (Nomura et al.,
1992) and D. Chogyoji (personal communication). The results showed
that the preference for glucose alone decreased from 0.35 ± 0.04
(n = 4) to 0.09 ± 0.07 (n = 4), whereas the
preference for glucose plus glycine increased from 0.65 ± 0.04
(n = 4) to 0.91 ± 0.07 (n = 4), as the concentration
of glycine was increased from 50 to 200 mM. When we used 200 mM fructose
instead of 400 mM glucose, taste enhancement was also observed in the presence
of 100 mM glycine. We also examined the effects of 100 mM serine (E) and 100
mM methionine (F) on feeding preference for glucose in C. japonicus.
However, serine and methionine did not affect the preference for glucose in
C. japonicus. As shown in Figure
4, 400 mM glucose induced feeding in several sets of 20 ants, but
<500 mM glycine did not induce feeding in any ants from the same
colony.
|
In Figure 5, the percentage
of ants showing a feeding response is plotted against the concentration of
glucose in the presence of 0 (A), 50 (B), 100 (C) or 200 mM glycine (D). The
concentrations of glucose at which 50 % of the ants showed the feeding
response, K1/2, were 490, 320, 220 and 100 mM in the
presence of 0, 50, 100 and 200 mM glycine, respectively. Statistically, the
distribution of the feeding threshold is regarded as a logarithmic normal
distribution, so that the value of K1/2 corresponds to the
median of the feeding threshold in a test population
(Hirakawa and Kijima, 1978).
Thus, glycine, even at concentrations that are ineffective in eliciting
feeding in the absence of glucose, significantly decreases the feeding
threshold of the ants to glucose, resulting in an increase of their preference
rate for glucose plus glycine, as shown in
Figure 3.
Effects of glycine on the sugar taste response of the chemosensillum
The electron micrograph of Figure 2A shows the dorsal view of the head of C. japonicus. Pairs of maxillary and the labial palps are seen (arrows in A). At high magnification (B), an arrow indicates a long type of sensillum located on the second segment of the labial palps. From this type of sensillum, we recorded sensory responses. The ants use the palps to share droplets of the DNO secretion.
In preliminary observations by scanning electron microscopy, we found several types of chemosensilla in the mouth-parts. Among them, an elongate taste sensillum on the second segment of the labial palps was the most reliable in retaining normal function under laboratory conditions. Therefore, we investigated the effects of glycine on the electrophysiological response of this sensillum.
First, we recorded sensory responses from this sensillum to 500, 100 and 10 mM NaCl (Figure 6). Stimulation with 500 mM NaCl induced impulses of large amplitude (A), but stimulation with 100 (B) or 10 mM NaCl (C) did not. These impulses were derived from a salt receptor cell. Stimulation with 100 or 10 mM NaCl induced impulses of small amplitude, which were derived from a water receptor cell (dots). As is characteristic of impulses from water receptor cells, the frequency decreased as the concentration of NaCl increased.
Since 100 mM NaCl only induced a small number of impulses from the water receptor cell, we used 100 mM NaCl to dissolve sugars and/or glycine for stimulus solutions. When the sensillum was stimulated with 400 mM glucose dissolved in 100 mM NaCl (D), a few impulses from the water receptor cell (dots) were also seen, but all other impulses differed in amplitude from impulses typically observed in either water or salt receptor cells. We concluded that these impulses of middle amplitude were derived from the sugar receptor cell. The absolute amplitudes of impulses of the salt (to 500 mM NaCl), the water (to 100 mM NaCl) and the sugar receptor cells (to 400 mM glucose in 100 mM NaCl) are listed in Table 1.
|
One mole of glycine dissolved in 100 mM NaCl induced, in <20 % of the trials, a small number of impulses from the sugar receptor cell (E, F). In most cases, however, glycine induced no impulses (G) or occasionally a small number of impulses of the water receptor cell (data not shown). In Figure 6, the impulses in A-C were recorded in the sensillum of one ant and those in D-G in different sensilla of another ant.
As shown in Figure 6, 1 M glycine typically did not induce electrophysiological responses from the sugar receptor cell (G), but, rarely, such responses were observed (E, F). Therefore, the sugar receptor cells in the sensilla on the labial palps may have responsiveness to glycine, but this response may be masked because the threshold for glycine is higher than that for glucose. Even if the ants have other types of taste organs sensitive to glycine than the sensilla examined here, their threshold for glycine would still be higher than that for glucose, because Figure 4B shows that only 15% of ants tested responded to 1 M glycine.
In order to investigate the concentration-dependent effect of glycine on the taste reception of glucose, sensory responses were recorded to 100 mM NaCl solutions of 400 mM glucose plus 0 (A), 50 (B), 100 (C) and 200 mM glycine (D), respectively (Figure 7). We observed that the impulses of the sugar receptor cell increased as the concentration of glycine increased, although the tested concentrations of glycine on its own did not excite the sugar receptor cell. We counted the number of impulses generated for 0.5 s (between black arrows in Figure 7) starting from 0.1 s after the beginning of stimulation. The same experiment was repeated in the same type of sensillum in four different animals, as shown in the diagram of Figure 8. The magnitudes of response to 400 mM glucose plus 100 or 200 mM glycine are significantly different from those to 400 mM glucose alone (Student's t-test, P < 0.05).
|
|
|
Figure 9A shows the glucose concentration-response curves in the absence and presence of 200 mM glycine. The relative magnitudes of response normalized to the response to 1 M glucose in the absence of glycine are plotted. We could not confirm whether the maximum response to glucose was changed by glycine or not, because the responses to concentrations of glucose >1 M were reduced in their amplitudes for unknown reasons and were, therefore, difficult to count. After stimulation with high concentrations of glucose, the responsiveness of the sensilla tended to be unstable, so that 32 sensilla examinations resulted in only three complete pairs of concentration-response curves comparable in the same sensilla. However, in the presence of 200 mM glycine (closed circles), the concentration-response curve appeared to shift to the left of the control curve in the absence of glycine (open circles). Since 200 mM glycine itself did not induce any impulses of the sugar receptor cell, it would appear that glycine has a synergistic rather than an additive effect on the response to glucose.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Role of amino acid in DNO secretion
Myrmecophilous lycaenid butterflies have a wide range of life cycles. Some
species, such as Jalmenus evagoras from Australia, feed on foliage
throughout their larval stages. The larvae and pupae offer attendant ants
nutritious secretions and, in return, the ants protect them against parasites
and predators (Pierce, 1984,
1987;
Pierce and Nash, 1999). Other
species, such as Maculinea arion, begin feeding on flowers, but after
the third instar, drop to the ground where they are picked up by ants and are
carried into the nest. They feed on the ant brood until pupation
(Hölldobler
and Wilson, 1994). Niphanda fusca, a Japanese
myrmecophilous lycaenid butterfly, is likewise a parasite of the attendant
ants, inducing their host ants, C. japonicus, to feed them
mouth-to-mouth by trophallaxis inside the nest. The larvae of N.
fusca are completely dependent upon their host ants for food.
Nevertheless, why should they simultaneously reward ants with secretions from
the DNO?
Although the DNO secretion may include other substances than sugars and
amino acids, studies on larval attractiveness to ants have focused on amino
acids, because the larvae of lycaenid species that feed on nitrogen-fixing
plants are much more likely to associate with ants than other lycaenid species
that feed on non-nitrogen-fixing plants
(Pierce, 1985; Fiedler,
1995,
1996). The DNO secretions of
some myrmecophilous lycaenid butterflies include particularly concentrated
amino acids; their content is often an order of magnitude more than that found
in most extrafloral nectaries or in honeydew of aphids
(Pierce and Young, 1986;
Pierce, 1987;
Fiedler and Maschwitz, 1989;
DeVries, 1991;
Fiedler and Saam, 1995). The
concentrated amino acids in the DNO secretions have been considered to be an
important nitrogen source for the attendant ants. Camponotus
japonicus cannot detect concentrations of glycine <500 mM (Figures
4 and
6), but ants attracted by the
DNO secretion can take up glycine, which could be a metabolically inexpensive
nitrogen source for them. Some social insects are known to use their larvae as
a kind of collective stomach. The adult workers feed complex proteins to the
brood. The brood digest these proteins and then regurgitate solutions rich in
free amino acids back to the workers (Hunt
et al., 1982). The workers can then use these amino acids
as precursors for the many chemicals used in communication and/or metabolic
processes. It is possible that the larvae of N. fusca use the
secretions from the DNO as a means of mimicking the ant brood.
Camponotus japonicus did not respond to glycine, serine or
methionine at the concentrations in the DNO secretion reported for N.
fusca (Nomura et al.,
1992) (Figure 4),
Jalmenus evagoras (Hunt et
al., 1982; Pierce,
1989) or Lysandra hispana
(Maschwitz et al.,
1975). However, C. japonicus obviously prefer
`glycine-flavored glucose', which mimics the sugar-amino acid mixture in the
DNO secretion of N. fusca, to plain glucose
(Figure 3A-C). Lysandra
hispana is a facultative and unspecific myrmecophile, whose ant visitors
cover a range of ant taxa (Fiedler,
1990). This contrasts sharply with the situation found in
Jalmenus evagoras, where the caterpillars do have a specific
relationship with only a small number of Iridomyrmex species and
records of associations with other ant species are from highly exceptional
situations (Eastwood and Fraser,
1999). Therefore, while in the specific myrmecophile Jalmenus
evagoras one might expect adaptation of secretions to the taste of the
specific host ant, such a specialization would be extremely unlikely in the
opportunistic myrmecophile Lysandra hispana. On the other hand,
different species of ants are known to be able to detect and prefer some amino
acids to others (Lanza, 1991;
Lanza et al., 1993).
The species of ants tending both Jalmenus evagoras and Lysandra
hispana do not, however, include C. japonicus and thus would be
expected to differ in their gustatory responses. Although we found that the
Japanese species of ants Formica japonica and C. obscripes,
which are not the attendant ants of N. fusca, would not prefer the
`glycine-flavored glucose' to glucose alone, further comparisons between
C. japonicus, the attendant ant of N. fusca, and the
attendant ants of Jalmenus evagoras or Lysandra hispana
should clarify the question of whether different kinds of amino acid increase
feeding preference in different species of ant.
Sweet-taste enhancement by amino acid in the ant
In mammals, synergistic taste-enhancing effects of `umami' substances are
well known. This is dramatically observed in synergistic interactions between
amino acids and nucleotides, whose tastes are classified into the fifth
fundamental taste, `umami' (Yamaguchi,
1967,
1991). `Umami' substances
result in increase in the palatability of foods and an improvement of appetite
in humans; similarly the taste-enhancing effects of glycine on the sweetness
of glucose result in an improvement of appetite in the ant, C.
japonicus (Figure 3).
Taste organs of insects have a unit structure containing a mechanoreceptor
and four taste receptor cells that differ in their functions. The salt and the
water receptor cells would not trigger feeding responses, as long as insects
appropriately maintain intrinsic salt and water balance. The so-called `fifth
cell' or `fourth taste receptor cell', for which adequate stimuli are not
conclusively determined, appears to respond to alkali halides such as CsCl
(Gillary, 1966), bitter
substances such as quinine (Liscia and
Solari, 2000) and lipophilic substances such as limonene (M.
Ozaki, unpublished data). These stimuli do not induce feeding responses, but
rather inhibit them and sometimes evoke aversion behavior in the blowfly
(Dethier, 1976;
Liscia and Solari, 2000;
Nakamura, et al.,
2000). Thus, even if the ants have this type of `fifth cell', it
is unlikely to be involved in glucose taste enhancement by glycine. The taste
receptor cell, excitation of which induces feeding response to sugars and
other nutritious substances such as amino acids, is traditionally called the
sugar receptor cell in insects. Because neither glucose
(Figure 6D) nor glycine at high
concentration (E, F) induced feeding response and evoked impulses of the sugar
receptor cell, it is suggested that the taste-enhancement mechanism by glycine
in C. japonicus is exclusively controlled by the sugar receptor
cell.
In insects, the responsiveness of the sugar receptor cells to amino acids
depends upon the species and their feeding style. The tsetse fly, for example,
can respond to various amino acids (Van
der Goes van Naters and Den Otter, 1998), but the fleshfly
responds to some (Shimada,
1975; Shimada and Isono,
1978). The blowfly responds to various sugars, but few amino acids
(Goldrich, 1973).
Camponotus japonicus may be more like the blowfly than the tsetse fly
or the fleshfly in the low responsiveness of its sugar receptor cell to
glycine (Figure 6E-G) and other
amino acids (data not shown). For enhancing the taste of glucose, however,
glycine need not act as a gustatory stimulus by itself, because even at lower
concentrations than the electrophysiological threshold, glycine enhanced the
taste of glucose (Figures 7 and
8). In other words, the glycine
receptor molecule would not mediate this taste enhancement.
The concentration-response curve shift observed in the presence of glycine
(Figure 9) might be explained
by an increase in stimulus affinity, the available number of receptor
molecules or the efficacy of some steps of the intracellular transduction
processes. As we have shown in the present study, however, 200 mM glycine
enhanced taste responses in C. japonicus, not only to glucose but
also to fructose (Figure 3).
Considering the multiple-site hypothesis, exemplified by the coexistence of
different receptor molecules for glucose and fructose on the sugar receptor
cell in the fleshfly (Shimada et
al., 1974) or the blowfly
(Hara, 1983), taste
enhancement by glycine in C. japonicus is presumed not to be a
receptor-molecule-specific event. In order to clarify the precise mechanism of
this taste enhancement, however, the target of glycine should be determined at
the molecular level after characterization of receptor molecules and
transduction mechanisms in the sugar receptor cell of the ants.
|
Acknowledgments |
|---|
We thank Mr H. Shimizu and Mr D. Chogyoji for helpful discussions and Professor T. Amakawa for kind permission to use his electrophysiological equipment. We also thank Professor N.E. Pierce for her critical reading of the paper and for helpful advice. This work was supported by a grant of Human Frontier Science Program to M.O.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Atsatt, P.R. (1981) Ant dependent oviposition by the mistletoe butterfly Ogyris amaryllis.Oecologia , 48,60 -63.
Dethier, V.G. (1976) The Hungry Fly. Harvard University Press, Cambridge, MA.
DeVries, P.J. (1991) Evolutionary and ecological patterns in myrmecophilous riodinid butterflies. In Huxley, C.R. and Cutler, D.F. (eds) Ant-Plant Interactions. Oxford University Press, Oxford, pp. 143-156.
Eastwood, R. and Fraser, A.M. (1999) Associations between lycaenid butterflies and ants in Australia.Austr. J. Ecol. , 24,503 -537.
Fiedler, K. (1990) European and north west African lycaenidae (Lepidoptera) and their associations with ants.J. Res. Lepid. , 28,239 -257.
Fiedler, K. (1995) Lycaenid butterflies and plants: is myrmecophily associated with particular hostplant preferences?Ethol. Ecol. Evol. , 7,107 -132.
Fiedler, K. (1996) Host-plant relationships of lycaenid butterflies: large-scale patterns, interactions with plant chemistry, and mutualism with ants. Ent. Exp. Appl.,80 , 259-267.
Fiedler, K. and Maschwitz, U. (1989) Functional analysis of the myrmecophilous relationships between ants (Hymenoptera: Formicidae) and lycaenids (Lepidoptera: Lycaenidae). 1. Release of food recruitment in ants by lycaenid larvae and pupae.Ethology , 80,71 -80.
Fiedler, K. and Saam, C. (1995) Ants benefit from attending facultatively myrmecophilous Lycaenidae caterpillars—evidence from a survival study.Oecologia , 104,316 -322.
Fiedler, K., Hölldobler, B. and Seufert, P. (1996) Butterflies and ants—the communicative domain. Experientia,52 , 14-24.
Gillary, H.L. (1966) Stimulation of the salt receptor of the blowfly. III. The alkali halides. J. Gen. Physiol., 50,359 -368.
Goldrich, N.R. (1973) Behavioral responses of Phormia regina (Meigen) to labellar stimulation with amino acids. J. Gen. Physiol., 61,74 -88.
Hara, M. (1983) Competition of polysaccharides with sugar for the pyranose and the furanose sites in the labellar sugar receptor cell of the blowfly, Phormia regina. J. Insect Physiol., 29,113 -118.
Henning, S.F. (1983) Biological groups within the Lycaenidae (Lepidoptera). J. Entomol. Soc. Southern Africa, 46,65 -85.
Hirakawa, Y. and Kijima, H. (1978) Behavioral analysis of glutathione receptor of Hydra. J. Comp. Physiol., 137,73 -81.
Hodgson, E.S., Lettvin, J.Y. and Roeder, K.D. (1955) Physiology of primary chemoreceptor unit.Science , 122,417 -418.
Hölldobler, B. and Wilson, E.O. (1990) The Ants. Harvard University Press, Cambridge, MA.
Hölldobler, B. and Wilson, E.O. (1994) Journey to the Ants. Harvard University Press Cambridge, MA.
Hunt, J.H., Baker, I. and Baker, H.G. (1982) Similarity of amino acids in nectar and larval saliva: the nutritional bases for trophallaxis in social wasps.Evolution , 36,1318 -1322.
Lanza, J. (1991) Response of fire ants (Formicidae, Solenopsis invicta and S. geminata) to artificial nectars with amino acids. Ecol. Entomol.,16 , 203-210.
Lanza, J., Vargo, E.L., Pulim, S. and Chang, Y.Z. (1993) Preferences of the fire ants Solenopsis invicta and S. geminata (Hymenoptera, Formicidae) for amino-acid and sugar components of extrafloral nectars. Environ. Entomol., 22,411 -417.
Liscia, A. and Solari, P. (2000) Bitter taste recognition in the blowfly: electrophysiological and behavioral evidence. Physiol. Behav.,70 , 61-65.
Maschwitz, V., Wust, M. and Schrian, I. (1975) Blaulingsraupen als Zuckerlieferauten für Ameisen. Oecologia,18 , 17-21.
Nagayama, F. (1950) Life cycle of Niphanda fusca. Ecol. Entomol., 3,9 -18.
Nakamura, K., Kakiuchi, M., Kawahara, Y., Ozaki, M. and Yamaoka, R. (2000) Associative memories between taste of sucrose and limonene or coumarin odors in the blowfly, Phormia regina.Jpn. J. Taste Smell Res. , 7,389 -390.
Nomura, K., Hirukawa, N., Yamaoka, R. and Imafuku, M. (1992) Problems on the symbiosis between the lycaenid butterfly larva, Niphanda fusca Shijimia and the ant Camponotus japonicus (1). J. Lepidopterol. Soc. Jpn,43 , 138-143.
Pierce, N.E. (1984) Amplified species diversity: a case study of an Australian lycaenid butterfly and its attendant ants. In Vane-Write, R.I. and Ackery, P.R. (eds), The Biology of Butterflies. Academic Press, London, pp.197 -200.
Pierce, N.E. (1985) Lycaenid butterflies and ants: selection for nitrogen-fixing and other protein-rich food plants. Am. Nat., 125,888 -895.
Pierce, N.E. (1987) The evolution and biogeography of associations between lycaenid butterflies and ants. In Harvey, P.H. and Partridge, L. (eds), Oxford Surveys in Evolutionary Biology, Vol. 4. Oxford University Press, Oxford, pp. 89-116.
Pierce, N.E. (1989) Butterfly-ant mutualism. In Grubb, P.J. and Whittaker, J. (eds), Towards a More Exact Ecology. Blackwell, Oxford, pp.299 -324.
Pierce, N.E. and Nash, D.R. (1999) The Imperial Blue, Jalmenus evagoras (Lycaenidae). In Kitching, R.I., Sheermeyer, E., Jones, R. and Pierce, N.E. (eds),Monographs on Australian Lepidoptera, Vol. 6. The Biology of Australian Butterflies . CSIRO Press, Sydney, pp.277 -316.
Pierce, N.E. and Young, W.R. (1986) Lycaenid butterflies and ants: two species stable equilibria in mutualistic, commensal, and parasitic interactions. Am. Nat., 128,216 -227.
Shimada, I. (1975) Two receptor sites and their relation to amino acid stimulation in the labellar sugar receptor of the fleshfly. J. Insect Physiol., 21,1675 -1680.
Shimada, I. and Isono, K. (1978) The specific receptor site for aliphatic carboxylate anion in the labellar sugar receptor of the fleshfly. J. Insect Physiol.,24 , 807-811.
Shimada, I., Shiraishi, A., Kijima, H. and Morita, H. (1974) Separation of two receptor sites in a single labellar sugar receptor of the fleshfly by treatment with p-chlormercuribenzoate. J. Insect Physiol.,20 , 605-621.
Tanimura, T., Isono, K., Takamura, K. and Shimada, I. (1982) Genetic dimorphism in the taste sensitivity to trehalose in Drosophila melanogaster. J. Comp. Physiol.,147 , 433-437.
Thomas, J.A., Elmes, G.W., Wardlaw, J.C. and Woyciechowski, M. (1989) Host specificity among Maculinea butterflies in Myrmica ant nests.Oecologia , 79,452 -457.
Van der Goes van Naters, W.M. and Den Otter, C.J. (1998) Amino acids as taste stimuli for tsetse flies.Physiol. Entomol. , 23,278 -284.
Wardlaw, J.C., Thomas, J.A. and Elmes, G.W. (2000) Do Maculinea rebeli caterpillars provide vestigial mutualistic benefits to ants when living as social parasites in Myrmica ant nests? Ent. Exp. Appl.,95 , 97-103.
Yamaguchi, S. (1967) The synergistic taste effect of monosodium glutamate and disodium 5' inosinate. J. Food Sci., 32,473 -478.
Yamaguchi, S. (1991) Basic properties of umami and effects on humans. Physiol. Behav.,49 , 833-841.
Accepted June 13, 2001
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||