& NEWS COMMENT
have been observed occasionally to capture smaller species7,8. Furthermore, the endemic Malagasy race praedicta belongs to a species that consists of two recognized races. The second race, morganii, is widely distributed over tropical Africa. In Madagascar, the guild of long-tongued hawkmoths is greater than in Africa (five species compared to three species with average tongue length >9 cm), and the Malagasy hawkmoth Coelonia solani, which is absent from Africa, almost matches praedicta in tongue length5. Thus, it seems more likely that praedictas long tongue has originated by a coevolutionary race between hawkmoths competing for nectar.
Deep flowers for long tongues
I
n 1862 Charles Darwin proposed that the evolution of flower depth and insect tongue length is driven by natural selection due to mutual gain from the plant animal interaction1. His partial evidence for such coevolution was the star orchid of Madagascar Angraecum sesquipedale, with a nectariferous flower spur 111/2 inches long. But no insect with a matching long tongue for nectar extraction was known. Darwin simply predicted the existence of a giant pollinator hawkmoth in Madagascar that was, tongue-wise, capable of handling the orchids extraordinary nectar position. His prediction was ridiculed by entomologists. In response, the star orchidhawkmoth interaction was discussed in greater detail in 1867 by Alfred Russel Wallace in his supportive paper Creation by law defending the theory of natural selection (Fig. 1)2. Fritz Mller (dubbed the Duke of Observation by Darwin), who had emigrated to Brazil in 1852, showed in 1873 via his brother Hermann that hawkmoths with tongues capable of sucking such deep-seated nectar existed3. The pollinator candidate hawkmoth of the star orchid of Madagascar was later recovered and described in 1903 under the historyassociative name Xanthopan morganii praedicta 4. However, while the relationship between shorter-spurred Angraecum orchid species and shorter-tongued hawkmoth species in natural populations in Madagascar has been verified5, Darwins famous predicted interaction between Angraecum sesquipedale and its pollinator hawkmoth, although continuously revolving in the scientific literature, has eluded empirical documentation. In a recent report in Botanica Acta, based on analyses of flower-visits by caged X. m. praedicta hawkmoths to flowers of A. sesquipedale, L.T. Wasserthal not only describes the pollination mechanism, but also challenges Darwins theoretical model for the flower depthtongue length evolution6. Wasserthals main contention is that praedictas impressive tongue length (up to 25 cm, Fig. 2) does not reflect coevolution with plants, as proposed by Darwin and Wallace, but evolution for avoidance of predators. In Wasserthals model, praedictas tongue elongation has primarily evolved to increase distance-keeping and sideways oscillating hovering against predators that ambush at visited shallow flowers of other nectariferous plant species. According to this model, the star orchid has been driven to great flower depth not by a coevolutionary race with
TREE vol. 13, no. 7 July 1998
praedicta but by a pollinator shift to it that is, one-sided plant adaptive evolution. Wasserthals report increases our knowledge about the possible natural interactions substantially. However, evidence for the proposed predator-avoidance hypothesis is not unequivocal. In particular, quantitative data showing that systematic predator-ambushing for large hawkmoths at flowers exists have never been reported. At most, single large mantids or spiders
Fig. 1. Natural selection at its famous extreme. Darwins predicted pollination of the star orchid of Madagascar as illustrated in Wallaces article2.
Copyright 1998, Elsevier Science Ltd. All rights reserved. 0169-5347/98/$19.00 PII: S0169-5347(98)01359-7
259
NEWS & COMMENT
long-tongued hawkmoths, such as the Neotropics9,10. Extreme tongue length is clearly not a shared derived character among the hawkmoth species4; it has arisen within several geographically separate lineages (one of which is X. morganii ). That evolution for the avoidance of (still unknown) predators would, in multiple, geographically separate lineages, result in extreme tongue elongation, and in turn in extreme flower depth guilds by pollinator shift, seems improbable. Finally, Wasserthals pollination mechanical data appear to contradict his own pollinator-shift hypothesis: pollen for transfer and adaptive spur evolution can only be carried at the basalmost portion of the tongue, whereas hawkmoths with tongues considerably than longer the flower depth (as in the case of a proposed shift to a suddenly great tongue length evolved for predator avoidance) receive distally attached pollen that becomes wasted. There is merely a mechanical constraint against great shifts in exploited tongue length. On balance, the evidence still suggests that the star orchid has been evolutionarily driven to great flower depths according to Darwins classical model of a coevolutionary race, whereby the hawkmoth visitor is compelled to insert its tongue into the very base, and thus to selectively pollinate, deeper flowers than its tongue reach11. The new hypothesis on the evolution of long tongues and deep flowers remains to be tested in the field. Wasserthals evidence for interaction in the wild between X. m. praedicta and A. sesquipedale was a single viscidium of a pollinarium discovered on the tongue of a field captured male introduced into the flight tent. What is still needed after 135 years is quantitative observational data of X. m. praedicta pollinating A. sesquipedale in natural populations. This magnificent orchid grows epiphytically in the coastal rainforest of Eastern Madagascar, a disappearing unique habitat. To explore Darwins prediction is indeed a race against time. Wasserthals study serves commendably to highlight the irreparable losses under way losses that will only be mitigated through cooperation with local researchers. L. Anders Nilsson
Dept of Systematic Botany, Uppsala University, Villavgen 6, S-752 36 Uppsala, Sweden ([e-mail])
References
1 Darwin, C. (1862) On the Various Contrivances by which British and Foreign Orchids are Fertilised by Insects, Murray 2 Wallace, A.R. (1867) Creation by law, Q. J. Sci. 4, 470488 3 Mller, H. (1873) Proboscis capable of sucking the nectar of Angraecum sesquipedale, Nature 8, 223 4 Rothschild, L.W. and Jordan, K. (1903) A revision of the lepidopterous family Sphingidae, Nov. Zool. 9 (Suppl.), 1972 5 Nilsson, L.A. et al. (1985) Monophily and pollination mechanisms in Angraecum arachnites Schlt. (Orchidaceae) in a guild of long-tongued hawk-moths (Sphingidae) in Madagascar, Biol. J. Linn. Soc. 26, 119 6 Wasserthal, L.T. (1997) The pollinators of the Malagasy star orchids Angraecum sesquipedale, A. sororium and A. compactum and the evolution of extremely long spurs by pollinator shift, Bot. Acta 110, 343359 7 Delf, B.H. and Harris, K.M. (1964) Notes on Nigerian hawkmoths, Nigerian Field 29, 150160 8 Wasserthal, L.T. (1996) Of hawkmoth species with long tongues, Rept. DFG 1/96, 2225 9 Silberbauer-Gottsberger, I. and Gottsberger, G. (1975) Ueber sphingophile Angiospermen Brasiliens, Plant Syst. Evol. 123, 157184 10 Haber, W.A. and Frankie, G.W. (1989) A tropical hawkmoth community: Costa Rican dry fo...