| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Reproductive Biology |
School of Biological and Conservation Sciences, University of KwaZulu-Natal, Private Bag X01, Scottsville, Pietermaritzburg 3209, South Africa
Received for publication September 4, 2004. Accepted for publication May 18, 2005.
ABSTRACT
Plants that lack floral rewards may nevertheless attract pollinators through mimetic resemblance to the flowers of co-occurring rewarding plants. We show how a deceptive orchid (Disa nivea) successfully exploits a reciprocally specialized mutualism between a nectar-producing plant (Zaluzianskya microsiphon) and its long-proboscid fly pollinator (Prosoeca ganglbaueri). Disa nivea is a rare southern African orchid known only from habitats that support large populations of Z. microsiphon, which it closely resembles in both general morphology and floral spectral reflectance. Significant covariation in floral traits of Z. microsiphon and D. nivea was detected among populations. Where mimics are uncommon, flies do not appear to discriminate between the flowers of the two species. Pollination success in D. nivea was much higher at a site with abundant Z. microsiphon plants than at a site where Z. microsiphon was rare. Exploitation of a highly specialized mutualism appears to demand a high degree of phenotypic resemblance to a rewarding model by a deceptive mimic, as exemplified by D. nivea. The majority of deceptive orchids, on the other hand, exploit relatively generalized pollination systems and thus require only a vague resemblance to rewarding plants in the community in order to attract pollinators.
Key Words: advergent evolution • Batesian mimicry • convergent evolution • deceptive • exploitation • long-tongued fly • mutualism • orchid • specialization
Although exploitation of mutualisms is widespread (e.g., Springer and Smith-Vaniz, 1972
; Janzen, 1975
; Anderson and Midgley, 2002
), we have only a sketchy understanding of the ecology and evolution of such interactions (Bronstein, 2001
). Conventional mutualism theory predicts that exploiters may have dire consequences for the evolution and persistence of obligate partnerships and could drive mutualisms to collapse (see Boucher et al., 1982
; Howe, 1984
; Soberon and Martinez del Rio, 1995). However, new conceptual and theoretical models (Bronstein, 2001
; Yu, 2001
; Yu et al., 2001
; Morris et al., 2003
) suggest that mutualists and exploiters can coexist. This indeed seems to be the case, as some of the most obligate specialized mutualisms coexist with exploiters and have probably done so for millenia (e.g., Pellmyr et al., 1996
; West et al., 1996
; Anderson and Midgley, 2002
).
Batesian mimicry in plants can be considered a type of mutualism exploitation (Bronstein, 2001
). The strategy of a Batesian mimic is to procure pollinators of other reward-producing plants without actually supplying any rewards of their own (see Fig. 1). The potentially negative effects apply to both the pollinators and the model(s) in the following ways (see Fig. 1): first, pollinators stand to lose fitness by wasting time and energy by visiting nectarless flowers; second, models may lose fitness when growing amongst mimics because pollinators may move away from areas where they have experienced nectarless flowers or develop preferences for other phenotypes associated with higher levels of rewards.
|
; Little, 1983
; Dafni, 1986
), but the number of cases that have been formally investigated remains few relative to protective mimicry in animals. It is only recently that biologists have begun to use both experimental and phylogenetic evidence to identify floral mimics (Dafni and Ivri, 1981
; Nilsson, 1983
; Johnson, 1994
, 2000
; Johnson et al., 2003a
). What has seldom been considered is whether specialization in the exploited mutualism has an influence on the evolution of a mimic. Bronstein (2001)
predicted that specialized mutualisms should most commonly be exploited by highly specialized exploiters, and we believe that this generality should also hold for floral Batesian mimicry. Conversely, deceptive species that exploit generalized pollination systems should possess a more generalized set of floral traits (cf. Nilsson, 1983
; Gumbert and Kunze, 2001
).
Relationships between plants and their pollinators are seldom reciprocally specialized, with most characterized by high levels of generalization on the part of pollinators and moderate generalization on the part of plants (Waser et al., 1996
; Johnson and Steiner, 2000
). The basis for this generalization and asymmetry is simply the need for pollinators to retain flexibility in their foraging, given that floral rewards from a given plant species are uncertain in both space and time (Waser et al., 1996
). Even plants seldom show specialization for a particular pollinator, as the risks of such specialization would normally outweigh the benefits (Aigner, 2001
).
In this study, we focus on the exploitation of an unusually specialized mutualism between Zaluzianskya microsiphon (Kuntze) K. Schum. (Scrophulariaceae) and the long-proboscid fly Prosoeca ganglbaueri (Nemestrinidae). Prosoeca ganglbaueri appears to be the exclusive pollinator of this species in the Drakensberg region (Johnson et al., 2002
; Figs. 2, 3). Furthermore, flowers of Z. microsiphon in the Drakensberg region appear to be the most important sources of nectar for P. ganglbaueri during a large proportion of its flight season. This is despite the fact that the fly has been recorded as a visitor to many different plant species across its range in South Africa (Goldblatt and Manning, 2000
).
|
In order to test the hypothesis that the orchid D. nivea is adapted to exploit the relationship between Z. microsiphon and the fly P. ganglbaueri, we sought answers to the following questions: (1) Does D. nivea depend on the long-proboscid fly P. ganglbaueri for its pollination? (2) Does D. nivea occur in association with Z. microsiphon? (3) Does D. nivea show similar floral dimensions and spectral reflectance to Z. microsiphon? (4) Is variation in floral morphology among populations of Z. microsiphon matched by variation in the floral morphology of D. nivea? (5) Is the pollination success of D. nivea higher at sites where Z. microsiphon is common than at sites where it is rare?
MATERIALS AND METHODS
Study species
Zaluzianskya microsiphon (Scrophulariaceae), the putative model in this system, is a nectar-producing species that is adapted for pollination by long-proboscid flies (Johnson et al., 2002
). Pollen of Z. microsiphon is deposited on the base of the flies' proboscis and on the hairs on the ventral parts of their heads (fig. 2 in Johnson et al., 2002
). Johnson et al. (2002)
showed that Z. microsiphon is self-incompatible and thus reliant on pollinator visits for seed set. The species occurs widely in the Drakensberg region (Fig. 7). In this region it has been observed flowering between January and March (B. Anderson and S. D. Johnson, personal observations; herbarium material in NU).
|
Prosoeca ganglbaueri (Nemestrinidae) is widespread throughout the Drakensberg region, where it has a flight period from January to March. Apart from Z. microsiphon, this fly is known to pollinate at least seven species in the Drakensberg region, and these form a guild characterized by cream or pink flowers with dilute nectar and often long corolla tubes (Johnson and Steiner, 1995
; Goldblatt and Manning, 2000
). The length of the proboscis of this fly varies from 19 to 42 mm between populations (Goldblatt and Manning, 2000
; B. Anderson and S. D. Johnson, unpublished data).
Study sites
The study was carried out between January and February 2002–2005 at all five sites where D. nivea is known to occur, namely Sani Pass, Bushmansnek, Ramatseliso's Gate, Qachasnek, and Matatiele in the Drakensberg region of southern Africa (Fig. 7). Zaluzianskya microsiphon was recorded from all of these sites. We recorded the approximate number of flowering individuals of Z. microsiphon and D. nivea within 100 x 100 m plots at each site. At all of the study sites, Z. microsiphon represented the only numerous rewarding plant belonging to the P. ganglbaueri guild (Goldblatt and Manning, 2000
). Approximately 30 flowering plants of Brunsvigia grandiflora Lindl. (Amaryllidaceae) were found 750 m from the Matatiele study site. Approximately 40 flowering plants of Hesperantha grandiflora G.J. Lewis (Iridaceae) were found 50 m from the Bushmansnek study site, and at Qachasnek, ca. 50 individuals of Gladiolus oppositiflorus Herb. were found intermingled with small patches of Z. microsiphon (Iridaceae). All of these species are also pollinated exclusively by P. ganglbaueri but bear pink flowers and are very different in appearance to Z. microsiphon and D. nivea.
Pollinator observations
Approximately 86 h of observations were carried out at the five sites over a period of 26 d to determine the pollinators of Z. microsiphon and D. nivea. Insects visiting the flowers and those seen near the Z. microsiphon and D. nivea populations were captured and examined for D. nivea pollinaria. Disa nivea pollinaria are easy to recognize because of their characteristic placement position on the proboscis and large size. In addition, no other large Disa species were flowering in close proximity. To obtain a measure of the diversity of other species whose flowers are visited by P. ganglbaueri, non-orchid pollen was removed from all of the flies captured at the study sites by rubbing their bodies with a small block of fuchsin gel, which was subsequently mounted on a slide for examination under a compound microscope (Beattie, 1971
). Pollen was identified using a reference collection made from pressed plants collected at the study sites. We stopped counting pollen grains when more than 1000 of any one plant species were encountered. Data on pollen loads were also obtained from flies captured by us at Witsieshoek, Royal Natal, and Sentinel (Fig. 7) and museum specimens collected from Giant's Castle, Garden Castle, and Ngeli (Fig. 7). Many of these specimens were captured while sunning themselves on rocks, therefore reducing bias towards a certain host plant.
Floral traits
To estimate the degree of morphological similarity between the two plant species, we measured inflorescence height, flower width, and flower depth of both Z. microsiphon and D. nivea at each site. Measurements were made in the field from one flower on each of 20 plants per species, except in the case of the Sani Pass site, where three pressed plants of each species were measured. No data for inflorescence height and flower width of either species were obtained for the Matatiele site. We used univariate regression to determine whether mean inflorescence and flower dimensions for D. nivea covary with those of Z. microsiphon across the study sites.
To obtain an objective estimate of similarity in flower color in Z. microsiphon and D. nivea, spectral reflectance over the UV-visible range (300–700 nm) was determined for five flowers of each species using an Ocean Optics (Dunedin, Florida, USA) S2000 spectrometer and Ocean Optics DT-mini deuterium tungsten halogen light source (200–1100 nm). Readings were taken through a fibre-optic reflection probe (UV/VIS 400 micron) held at 45° and about 5 mm from the surface of the petal.
Pollinator morphology
To establish whether there is covariation between flower and pollinator dimensions, we measured the length of the proboscis of flies captured at each of the study sites. The proboscis of P. ganglbaueri can be telescoped in and out by a combination of musculature action and haemolymphal pressure in a manner similar to that of bee-flies (Grimaldi, 1988
; Gilbert and Jervis, 1998). The proboscis of nemestrinid flies consists of two parts: a membranous, basal part that can extend and contract and a much longer anterior section of fixed length. We measured only the anterior section of the proboscis in order to standardize all measurements. Then, using 10 fresh flies with proboscis lengths varying between 21.7 and 42.4 mm, we made a crude correction value to calculate the length of the fully extended proboscis. On each of these flies, we measured the length of the fixed anterior section and then manually stretched out the proboscis to ascertain the fully extended length. The correction value obtained was 20.25 ± 2.06%, which is similar to the value of 28% reported for bee-flies by Grimaldi (1988)
. This conversion factor was used to obtain a crude functional length of fly proboscides in this paper. Univariate regression was used to determine the relationship between flower depth of the two plant species and the mean proboscis length of P. ganglbaueri across the five study sites.
Pollination success
The pollination success of Z. microsiphon was estimated from the proportion of flowers that set fruit and from the mean number of seeds in fruits. A single flower capsule was haphazardly chosen from each of 26 plants at Ramatseliso's Gate and 23 plants at Qachasnek. Pollination success for D. nivea was estimated more directly by examining a single haphazardly selected flower on each of 59 plants at Ramatseliso's Gate and 34 plants at Qachasnek and recording whether or not pollinaria had been removed and whether or not there was pollen present on the stigma. This was done at two sites: Ramatseliso's Gate, where there were approximately 10 times as many Z. microsiphon plants as there were orchids, and Qachasnek, where the number of individuals of Z. microsiphon and D. nivea was approximately the same. These two sites were chosen because they are matched in terms of altitude and aspect, yet differed in the proportions of the two species.
RESULTS
Pollination observations
At four of the five study sites, P. ganglbaueri was the only insect species observed to visit flowers of Z. microsiphon, and all captured flies carried conspicuous pollen loads of Z. microsiphon on the ventral surface of the head and proboscis (Fig. 2; Table 1). At Bushmansneck we captured an additional, unidentified nemistrinid pollinating Z. microsiphon.
|
|
Floral traits
The mean widths of flowers of D. nivea and Z. microsiphon were similar at each site and had a significant positive relationship across four populations (Fig. 8). Inflorescence height, on the other hand, varied greatly within each site for both Z. microsiphon and D. nivea, and there was considerable overlap both within and between sites (Fig. 8). No significant relationship was detected for mean inflorescence height of the two species (Fig. 8). Mean flower depth for the two plant species was similar at each site and showed a highly significant relationship (Fig. 8). Flower depth varied significantly among populations for both plant species (Z. microsiphon, F = 552, P < 0.01; D. nivea, F = 201, P < 0.01; Fig. 9).
|
|
|
Pollination success
Measures of both seed set and fruit set were almost equally high at both sites for Z. microsiphon (Fig. 11). In contrast, male and female pollination success of D. nivea was very much lower at Qachasnek than at the Ramatseliso's gate site (Fig. 11). Zaluzianskya microsiphon plants greatly outnumbered D. nivea plants at the Ramatseliso's gate site (Table 1; Fig. 11).
|
Is Disa nivea a floral mimic?
Disa nivea satisfies many of the criteria needed to characterize a floral Batesian mimic (see Dafni and Ivri, 1981
; Ackerman, 1986
; Johnson, 1994
). It has a similar flowering phenology to Z. microsiphon and occurs in close association with its model (and mostly at lower densities). The plants also share a common pollinator, the long-proboscid fly P. ganglbaueri. These pollinators were often captured with several pollinaria of D. nivea on the base of their proboscides, indicating that they had visited the orchid on more than one occasion. This is probably due to the very similar spectral reflectance of flowers of the two plants and also similarities between other morphological traits such as inflorescence height and floral dimensions. Not only are the widths of D. nivea and Z. microsiphon very similar, but they were also found to covary across sites. Neither species has a floral scent detectable to humans. Finally, our data also suggest that when the mimic outnumbers the model, then the fitness of the mimic is likely to decrease, although this is not conclusive as it is based on comparison of just two populations. Although differences in pollination success of the orchid between two of the sites may simply be due to difference in pollinator abundance, vis-à-vis the magnet species effect (Johnson et al., 2003b
), flowers of Z. microsiphon were well pollinated at both sites, as adjudged by high seed set in this self-incompatible species. Our interpretation is that flies occurred at both sites, but discriminated between D. nivea and Z. microsiphon only at sites where the mimic was common and frequently encountered.
How specialized is the exploited mutualism?
Although not as specialized as yucca or fig mutualisms, the mutualism between Z. microsiphon and P. ganglbaueri is unusually reciprocally specialized for a pollination mutualism (Waser et al., 1996
). At the population level our results (and those of Johnson et al., 2002
) indicate that Z. microsiphon is pollinated by a single fly pollinator species at most study sites in the Drakensberg region. Furthermore, the species is self-incompatible and thus unable to set seed in the absence of insect visits (Johnson et al., 2002
).
Although P. ganglbaueri pollinates a large number of plants across its range (Goldblatt and Manning, 2000
), we found no more than two rewarding guild members co-flowering at any of the study sites. Of these plants, Z. microsiphon was by far the most numerous, forming extensive populations of several thousand plants. In contrast, the other plants of this guild were comparatively rare and localized. These observations were corroborated by the finding that Z. microsiphon pollen made up the great majority of the pollen loads from flies at each site. In comparison, large pollen loads from other flowers were localized (e.g., G. oppositiflorus, one site) and loads from other flowers were insignificant. This suggests that at this time of year, flies depend heavily on Z. microsiphon for their nectar requirements and that Z. microsiphon is by far the most important food plant for P. ganglbaueri over much of its range in the Drakensberg region. Thus, our results are similar to Fox and Morrow (1981)
who found that herbivorous insects may function as generalists over their entire range, but may be highly specialized at the local level. The data also supports the prediction of Thompson (1994)
that local specialization may be important in mutualisms as well as parasitisms and plant–herbivore interactions.
How specialized is the exploiter?
Disa nivea is a specialized exploiter on two different levels. First, it appears to exploit a single pollinator species across most of its range. Second, it appears to exploit or mimic a single plant species across its entire range, tracking corolla length and flower width of Z. microsiphon across several sites and having a very close color match. Zaluzianskya microsiphon is the only nectariferous species in the P. ganglbaueri guild that bears any resemblance to D. nivea (these are the only two white/cream-colored flowers in a predominantly pink guild) and is the only species consistently sympatric with D. nivea. Thus Bronstein's (2001
, p. 279) prediction that "highly specialized mutualisms should most commonly be exploited by highly specialized exploiters" appears to hold.
Is the resemblance between mimic and model the result of advergent or convergent evolution?
Johnson et al. (2003a)
pointed out that floral similarities among unrelated plant species that share pollinators can arise through two different evolutionary pathways. Advergent evolution describes the scenario when phenotypes of some plant species remain unchanged, while selection acts on the phenotypes of a second group, bringing about a resemblance to the first group. Convergent evolution is the more familiar scenario when a common selective pressure is applied to all members of the guild bringing about an overall resemblance. This distinction serves to draw attention to two very different origins of the selection pressures applied to floral traits. Convergent evolution is based primarily on innate pollinator preference and was traditionally believed to underlie much of the evolution of floral syndromes. Advergent evolution on the other hand is based primarily on conditioned pollinator preferences. There has been an increasing realization that pollinators select flowers more on the basis of conditioned preferences than innate preferences (Chittka and Thomson, 2001
) and thus that advergent evolution may be more common in nature than was previously thought to be the case. Advergent evolution is exemplified by floral mimicry because the resemblance between species is entirely due to evolutionary modification of the mimic and not the model (Johnson, 1994
, 2000
; Johnson et al., 2003a
).
While our results show clearly that the mutualism between Z. microsiphon and P. ganglbaueri is exploited ecologically by D. nivea, it is more difficult to demonstrate that the floral traits of the orchid have evolved in response to foraging preferences of flies conditioned by feeding on Z. microsiphon (i.e., advergent evolution, sensu Johnson et al., 2003a
), as opposed to simple innate preferences of flies, in which case the similarity could simply be due to convergent evolution within a floral guild.
We examined three morphological traits (flower color, flower width, and flower depth), all of which were found to be similar in D. nivea and Z. microsiphon. We argue that the strikingly similar reflectance spectra of D. nivea and Z. microsiphon is likely the result of advergent (assymetrical) evolution. If cream colors evolved in response to innate preferences by P. ganglbaueri, then we would expect convergent evolution for this color among all the approximately 20 plant species that depend primarily on this fly for their pollination. However, Z. microsiphon is the only rewarding species in the guild with cream flowers, almost all of the others having pink flowers (Goldblatt and Manning, 2000
). A similar argument could be advanced for flower size, as P. ganglbaueri pollinates flowers with a range of sizes, including several with massive flowers such as those of Brunsvigia grandiflora (Amaryllidaceae). We interpret the significant relationship between mean flower width of D. nivea and Z. microsiphon in different populations (Fig. 8) as trait-tracking of the latter taxon by the former.
In contrast, the significant relationship between mean flower depth of the two species is unlikely to be due to advergent evolution because this is unlikely to be a trait that affects conditioned foraging preferences by pollinators. Instead, this relationship is likely to reflect independent adaptation to pollinator proboscis length in the two plant species, as indicated by the significant correlations found between flower depth and average proboscis length of the flies.
Conclusion
This study demonstrates that D. nivea possesses a remarkably specialized pollination system based on the exploitation of the mutualism between Z. microsiphon and its pollinator P. ganglbaueri. The success of this exploitation by D. nivea is based on its apparent adaptive resemblance to the flowers of Z. microsiphon. Disa nivea probably represents one extreme on the continuum from deceptive orchids with a specialized phenotype that mimics a single plant species (cf. Nilsson, 1983
; Johnson, 1994
, 2000
; Johnson et al., 2003a
) to those with a highly generalized phenotype approximating that of "typical" plants visited by generalist pollinators (cf. Nilsson, 1983
; Johnson et al., 2003b
). It is unlikely to be a coincidence that well-developed floral mimicry in orchids is associated with exploitation of highly specialized pollination systems, such as those involving oligolectic bees and long-proboscid flies (Nilsson, 1983
; Johnson, 1994
, 2000
).
FOOTNOTES
1 The authors thank the following organizations for providing financial support for this study: the University of KwaZulu-Natal (postdoctoral fellowship to B. A.) and the National Research Foundation (doctoral scholarship to C. C., research grant to S. D. J.). We also thank Jana Jersakova for help with fieldwork and David Barraclough for identification of insects. 
2 Author for correspondence (e-mail: johnsonsd{at}ukzn.ac.za
)
LITERATURE CITED
Ackerman J. D. 1986 Mechanisms and evolution of food-deceptive pollination systems in orchids. Lindleyana 1: 108-113
Aigner P. A. 2001 Optimality modeling and fitness trade-offs: when should plants become pollinator specialists?. Oikos 95: 177-184[CrossRef][ISI]
Anderson B. J. J. Midgley 2002 It takes two to tango but three is a tangle: mutualists and cheaters on the carnivorous plant Roridula. Oecologia 132: 369-373[CrossRef][ISI]
Beattie A. J. 1971 A technique for the study of insect-borne pollen. Pan-Pacific Entomologist 47: 82
Bronstein J. L. 2001 The exploitation of mutualisms. Ecology Letters 4: 277-287
Boucher D. H. S. James K. H. Keeler 1982 The ecology of mutualism. Annual Review of Ecology and Systematics 13: 315-347
Chittka L. J. D. Thompson 2001 Cognitive ecology of pollination. Cambridge University Press, Cambridge, UK
Dafni A. 1986 Floral mimicry—mutualism and unidirectional exploitation of insects by plants. In T. R. E. Southwood and B. E. Juniper [eds.], The plant surface and insects, 85–94. Edward Arnold, London, UK
Dafni A. Y. Ivri 1981 Floral mimicry between Orchis israelitica Baumann and Dafni (Orchidaceae) and Bellavalia flexuosa Boiss (Liliaceae). Oecologia 49: 229-232[CrossRef][ISI]
Fox L. R. P. A. Morrow 1981 Specialization: species property or local phenomenon?. Nature 211: 887-893[CrossRef]
Gilbert F. M. Jervis 1988 Functional, evolutionary and ecological aspects of feeding-related mouthpart specializations in parasitoid flies. Biological Journal of the Linnean Society 63: 495-535
Goldblatt P. J. C. Manning 2000 The long-proboscid fly pollination system in southern Africa. Annals of the Missouri Botanical Garden 87: 146-170[CrossRef][ISI]
Grimaldi D. 1988 Bee flies and bluets: Bombylius (Diptera: Bombyliidae) flower-constant on the distylous species, Hedyotis caerulea (Rubiaceae), and the manner of foraging. Journal of Natural History 22: 110
Gumbert A. J. Kunze 2001 Colour similarity to rewarding model plants affects pollination of food deceptive orchid, Orchis boryi. Biological Journal of the Linnean Society 72: 419-433[CrossRef]
Howe H. F. 1984 Constraints on the evolution of mutualisms. American Naturalist 123: 764-777[CrossRef][ISI]
Janzen D. H. 1975 Pseudomyrmex nigropilosa: a parasite of a mutualism. Science 188: 936-938
Johnson S. D. 1994 Evidence for Batesian mimicry in a butterfly-pollinated orchid. Biological Journal of the Linnean Society 53: 91-104
Johnson S. D. 2000 Batesian mimicry in the non-rewarding orchid Disa pulchra, and its consequences for pollinator behaviour. Biological Journal of the Linnean Society 71: 119-132
Johnson S. D. K. E. Steiner 1995 Long-proboscid fly pollination of two orchids in the Cape Drakensberg mountains, South Africa. Plant Systematics and Evolution 195: 169-175[CrossRef][ISI]
Johnson S. D. K. E. Steiner 2000 Generalization versus specialization in plant pollination systems. Trends in Ecology and Evolution 15: 140-143
Johnson S. D. T. J. Edwards C. Carbutt C. Potgieter 2002 Specialization for hawkmoth and long-proboscid fly pollination in Zaluzianskya section Nycterinia (Scrophulariaceae). Botanical Journal of the Linnean Society 138: 17-27[CrossRef]
Johnson S. D. R. Alexandersson H. P. Linder 2003a Phylogenetic and experimental evidence for floral mimicry in a guild of fly-pollinated plants. Biological Journal of the Linnean Society 80: 289-304[CrossRef]
Johnson S. D. C. I. Peter L. A. Nilsson J. Ågren 2003b Pollination success in a deceptive orchid is enhanced by co-occurring rewarding "magnet" plants. Ecology 84: 2919-2927[CrossRef][ISI]
Little R. J. 1983 A review of food deception mimicries with comments on floral mutualism. In C. E. Jones and R. J. Little [eds.], Handbook of experimental pollination biology, 294–309. Van Nostrand Reinhold, New York, New York, USA
Morris W. F. J. L. Bronstein W. G. Wilson 2003 Three-way coexistence in obligate mutualist–exploiter interactions: the potential role of competition. American Naturalist 161: 861-875
Nilsson L. A. 1983 Mimesis of bellflower (Campanula) by the red helleborine orchid Cephalanthera rubra. Nature 305: 799-800[CrossRef]
Pellmyr O. J. Leebens-Mack C. J. Huth 1996 Non-mutualistic yucca moths and their evolutionary consequences. Nature 380: 155-156[CrossRef][Medline]
Soberon Mainero J. C. Martinez del Rio 1985 Cheating and taking advantage in mutualistic associations. In D. H. Boucher [ed.], The biology of mutualism, 192–216. Croom Helm, London, UK
Springer V. G. W. F. Smith-Vaniz 1972 Mimetic relationships involving the fishes of the family Bleniidae. Smithsonian Contributions to Zoology 112: 1-36
Thompson J. N. 1994 The coevolutionary process. University of Chicago Press, Chicago, Illinois, USA
Yu D. W. 2001 Parasites of mutualisms. Biological Journal of the Linnean Society 72: 529-546[CrossRef]
Yu D. W. H. B. Wilson N. E. Pierce 2001 An empirical model of species coexistence in a spatially structured environment. Ecology 82: 1761-1771[CrossRef][ISI]
Waser N. M. L. Chittka M. V. Price N. M. Williams J. Ollerton 1996 Generalization in pollination systems, and why it matters. Ecology 77: 1043-1060[CrossRef]
West S. A. E. A. Herre D. M. Windsor P. R. S. Green 1996 The ecology and evolution of the New World non-pollinating wasp communities. Journal of Biogeography 23: 447-458[CrossRef][ISI]
Wiens D. 1978 Mimicry in plants. Evolutionary Biology 11: 365-403
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
