For many species of conservation significance, multiple factors limit reproduction. In a new study published in AoB PLANTS, Walsh et al. examined the contribution of plant height, number of flowers, number of stems, as well as the joint impacts of mutualists and antagonists on the pollination biology and seed production of the imperiled, deceptive orchid, Cypripedium candidum. They found flowering stem height to be the only morphological feature significant in reproduction, with taller flowering stems simultaneously receiving increased pollination and decreased seed predation. Furthermore they found decreased seed mass in individuals subjected to hand-self pollination treatments. Their results may help explain the factors limiting seed production in other Cypripedium and further emphasize the importance of management in orchid conservation.
Understanding the species composition of pollen on pollinators has applications in agriculture, conservation and evolutionary biology, but current identification methods cannot always discriminate taxa at the species level. Kron et al. test the use of flow cytometry to characterize pollen loads from individual bees, using DNA content as a species marker, and find that they are able to quickly measure DNA contents for nuclei from hundreds to thousands of pollen grains per bee. They observe differences in pollen load diversity between bumble-bees and honey-bees and find evidence of between-cytotype pollinator movement in a population of Solidago. This technique provides a new tool to complement other methods for examining pollinator behaviour.
A key innovation in the evolution of plants was the origin of the hermaphroditic flower, where both male and female sexual functions occur in the same complex structure. However, this innovation created a significant problem: sexual conflict, in which the function of one sex is compromised by the proximity and function of the other. This led to a further fundamental challenge in the function of animal-pollinated, hermaphroditic flowers: minimizing such sexual conflict while still enabling the male and female fertile parts to contact pollinators in the same place. Two solutions to sexual conflict have been explored evolutionarily by plants: (1) spatial separation of fertile parts (herkogamy) and (2) temporal separation of sexual functions (dichogamy).
To evaluate the effect of partial dichogamy and movement herkogamy on pollination accuracy in ‘generalist’ flowers (flowers pollinated by a variety of animal species), a recent paper in Annals of Botany investigates Parnassia epunctulata, a plant with open, white flowers, from subalpine meadows. The stamens of this species show a remarkable pattern of repositioning, and dehisce one by one over several days before the female phase. This feature permitted the authors to examine whether the anthers and stigma are positioned accurately, facilitating pollen removal and receipt.
The open flowers were visited by a variety of pollinators, most of which were flies. Seed set was pollinator-dependent (bagged flowers set almost no seeds) and pollen-limited (manual pollination increased seed set over open pollination). Analyses of adaptive accuracy showed that coordinated stamen movements and style elongation (movement herkogamy) dramatically increased pollination accuracy. Specifically, dehiscing anthers and receptive stigmas were positioned accurately in the vertical and horizontal planes in relation to the opposite sexual structure and pollinator position. In contrast, the spatial correspondence between anthers and stigma was dramatically lower before the anthers dehisced and after stamens bent outwards, as well as before and after the period of stigmatic receptivity. This shows for the first time that a combination of movement herkogamy and dichogamy can maintain high pollination accuracy in flowers with generalized pollination. Staggered pollen and stigma presentation with spatial correspondence can both reduce sexual interference and improve pollination accuracy.
Armbruster, W.S., Corbet, S.A., Vey, A.J., Liu, S.J., & Huang, S.Q. (2013) In the right place at the right time: Parnassia resolves the herkogamy dilemma by accurate repositioning of stamens and stigmas. Annals of Botany, 113 (1): 97-103.
A significant proportion of orchids in the subtribe Oncidiinae produce floral oil as a food reward that attracts specialized bee pollinators. This oil is produced either by glands (epithelial elaiophores) or by tufts of secretory hairs (trichomal elaiophores). Although the structure of epithelial elaiophores has been well documented, trichomal elaiophores are less common and have not received as much attention.
The flowers of Lockhartia are 5–30 mm in length and lack fragrance perceptible to humans. Oil secretion by flowers of Lockhartia was first reported by Silvera (2002), but the morphology and anatomy of their elaiophores have not previously been studied in detail. A recent paper in Annals of Botany surveys the flowers of 16 species of Lockhartia and shows that all have elaiophores (oil glands) of the trichomal type.
Specialized hairs on the legs or abdomen (but not the mouthparts) of oil-gathering bees are used to collect oils, and the latter are then used as food for larvae. Pollinaria of Lockhartia are small (typically 0·7–1·3 mm long) and their attachment to the bodies of bees has not been reported. This may be due to the fact that the thin stipe collapses upon drying and this obfuscates identification of the pollinarium to generic level. The situation is further exacerbated by the fast-flying and extremely timid nature of oil-collecting bees. As a result, they are much more difficult to capture or observe from short distances than male euglossine bees, for which an abundance of observational data exists.
Blanco, M. A., Davies, K. L., Stpiczyńska, M., Carlsward, B. S., Ionta, G. M., & Gerlach, G. (2013). Floral elaiophores in Lockhartia Hook. (Orchidaceae: Oncidiinae): their distribution, diversity and anatomy. Annals of Botany, 112(9), 1775-1791.
Pat Willmer. 2011. Princeton University Press. £65. pp. 832.
Any text book that tries to assess and summarise the whole of a multidisciplinary research field such as pollination ecology and floral biology is required to be four things: (1) comprehensive in its scope; (2) up to date in its coverage of the literature; (3) accurate in its assessment of the current state of the field; and (4) authoritative in the conclusions it presents.
This volume by Professor Pat Willmer of the University of St Andrews certainly ticks the first box. It’s a huge book, and covers everything relating to the evolution of flower attraction and reward systems, ecological interactions with pollinators, biochemistry, physiology, agriculture and conservation; all in 29 chapters split into three sections, with 87 pages of references. The literature extends to 2010, which is impressive for a book published in 2011 (though see my comments below about completeness of the literature). Specialist terms are highlighted in bold to direct the reader to the glossary at the back, a useful device even if there are a few inaccuracies, which I’ll mention later.
So far so good, and the author is to be congratulated on putting together such a comprehensive, not to mention timely, single-author book. It’s clearly the summation of a career devoted to studying pollinators and flowers, and the author’s passion for her subject is apparent throughout.
However when we come to points 3 and 4, things are less straightforward. There are some issues with accuracy that are troubling in a book aimed at newcomers to the field as well as established researchers. To give just a few examples:
- on p.18 we are told that asclepiads have “one stamen” (they have five); on p.169 and in the glossary that asclepiad pollinia are the pollen grains from one anther (they are the contents of half an anther); and on p.170 that the pollinaria are “glued” to pollinators (they actually clip on).
- in the glossary, tree ferns are referred to as “cycads”, an error that is repeated on p.89.
- on p.88 there is a statement suggesting that tree fern spores were dispersed by “animal fur” 300 million years ago, long before the evolution of mammals, and that this (and dispersal of spores of fungi and mosses) is the equivalent of pollination: it is not, it equates to seed dispersal.
These are troubling errors of basic botany that are forgivable in an early draft of the book (everyone makes mistakes) but not in the final published version, after it’s been read, reviewed, checked and edited. If the book goes to a second edition I hope that these (and other) mistakes will be fixed. But they do hint at a fundamental problem with a book (and a field) as large and complex as this: a single author is arguably unlikely to be able to do justice to all of the subject matter.
There are parts of the book where it is unclear (to me at least) what the author is actually saying. For example, on p.96 there is a graph which, it is suggested, demonstrates that pollination by animals is “technically uncommon when assessed in terms of the numbers of broad taxonomic groups that use it”, though the legend to the figure claims that “most orders of plants have no families” that possess wind pollination. This is confusing: what is to be concluded by someone new to the field? Is animal pollination common or rare? Likewise, on p.91 we are told that the “first angiosperms… would probably have had their pollen moved mainly by wind…”, but then on p.92 that “an element of insect pollination could be regarded as almost ancestral”. Which is correct?
There are other aspects to the book that are simply out of date; for example the linear, rather deterministic schemes set out in Figures 4.6 and 4.8 showing that Cretaceous flowers were open and radially symmetrical, and only later evolved into complex, bilateral flowers in the Tertiary, ignores fossil discoveries showing that orchids evolved in the Cretaceous (Ramírez et al., 2007). Likewise, discussion of “counterproductive” crypsis in flowers (p.124) neglects recent findings of cryptic, wasp-pollinated plants in South Africa (e.g. Shuttleworth & Johnson, 2009).
There is a theme emerging here: some of the botany that the book presents is inaccurate, confused or out-dated. Fortunately the zoological aspects of the book are much better, as one might hope from a Professor of Zoology.
The final criterion, that the book should be “authoritative in the conclusions it presents”, is however, in my view, the main weakness of this volume. The author is unhappy with recent developments in the field, particularly as they relate to community-scale assessments of plant–pollinator interactions, in terms of network analyses and predictive utility of pollination syndromes. Clearly Professor Willmer is on a mission to rebalance what she perceives as failings within some of the current trends in studying pollination. A book review is not the place for a technical dissection of the author’s arguments, which is best left to the peer-reviewed literature (though I would argue that that’s also the place to present some of the criticisms the author introduces, rather than into a text book such as this). I could focus the whole of this review on these topics because: (a) they take up a large proportion of the book, about one-third of the text pages; and (b) they are highlighted on the cover as being one of the main contributions of the book; specifically, that the author provides a critique of previous work that does not distinguish between “casual visitors and true pollinators” that can in turn result in “misleading conclusions about flower evolution and animal-flower mutualism”. Unfortunately her targets are straw men, and one – I believe quite telling – example will suffice.
On p.447 there is a criticism of the use by Waser et al. (1996) of Charles Robertson’s historical data set, and specifically that the analyses they present “…did not distinguish visitors from pollinators even though Robertson’s database did include information on this”. However Waser et al. clearly state (p.1045 of their paper) that only pollinators were included in the analyses, not all flower visitors, and that “visitation is not a synonym for pollination… non-pollinating visitors are excluded (as in Robertson 1928)” (p.1048).
Why should Professor Willmer make a statement to the contrary? Evidently she wishes to impress upon her readers that (in her opinion) there are fundamental problems in current approaches to studying pollination at a community level. But even if that were the case (and I don’t believe it is) misrepresenting previous studies to suit an argument is poor scholarship at best.
Regardless of whether some of her criticism is well founded, the author does not seem to appreciate that plant–flower visitor interaction networks are ecologically important regardless of whether or not a flower visitor acts as a pollinator. More fundamentally, true pollination networks possess similar attributes to flower visitor networks, for example a nested pattern of interactions, and arguments about level of generalisation of species are a matter of scale, not category (Ollerton et al., 2003).
At the end of her Preface, Professor Willmer reveals to us quite a lot about her personal attitude to research when she states that some readers might find her approach “too traditional” in an “era where ecological modelers [might be claimed to] have more to tell us than old-style field workers”. What the author fails to appreciate is that this is a grossly false dichotomy and that most of the pollination ecologists who have embraced new analytical methodologies for understanding plant–pollinator interactions are also “old-style field workers” with considerable experience of studying the ecology of flowers and their pollinators beyond the computer screen.
In summary this is a book that, for all its good qualities of comprehensiveness and (mostly) up to date coverage, should be read with caution: parts of it are neither as accurate nor as authorative as the field of pollination and floral ecology deserves.
Ollerton J, Johnson SD, Cranmer L, Kellie, S. 2003. The pollination ecology of an assemblage of grassland asclepiads in South Africa. Annals of Botany 92: 807-834.
Ramírez SR, Gravendeel B, Singer RB, Marshall CR, Pierce NE. 2007. Dating the origin of the Orchidaceae from a fossil orchid with its pollinator. Nature 448: 1042-1045.
Shuttleworth A, Johnson SD. 2009. The importance of scent and nectar filters in a specialized wasp-pollination system. Functional Ecology 23: 931-940.
Waser NM, Chittka L, Price MV, Williams N, Ollerton J. 1996. Generalization in pollination systems, and why it matters. Ecology 77: 1043-1060.
Plants surrounded by individuals of other co-flowering species may attract more pollinators but can suffer a reproductive cost from interspecific pollen transfer. Yang et al. compare pollination and reproduction in Pedicularis densispica (lousewort) when occurring alone or together with co-flowering Astragalus pastorius. They find that mixed populations attract many more nectar-seeking bumble-bees, which move frequently between the species. However, differences in floral architecture mean that P. densispica is pollinated via the dorsum of the bees whilst A. pastorius receives pollen via the abdomen, thus avoiding interspecific transfer. The overall result is that co-flowering yields more seeds that are heavier and have higher germinability than in pure populations of P. densispica.
The role of flower specialization in plant speciation and evolution remains controversial. Fernández-Mazuecos et al. use a time-calibrated phylogeny in conjunction with morphometric analysis to study bifid toadflaxes (Linaria sect. Versicolores), which have highly specialized corollas. They determine that a restrictive character state (narrow corolla tube) is reconstructed in the most-recent common ancestor. After its early loss in the most species-rich clade, this character state has been convergently reacquired in multiple lineages of this clade in recent times, yet it seems to have exerted a negative influence on diversification rates. The results suggest that opposing individual-level and species-level selection pressures may have driven the evolution of pollinator-restrictive traits in the bifid toadflaxes.
The magnitude of pollen limitation is highly variable among habitats, species and differing plant communities. Alonso et al. analyse natural variability in pollen receipt and tube formation, and compare pollen quality and quantity between co-flowering endemics and non-endemics at three biodiversity hotspots in Andalusia, California and Yucatan. They find that only a combination of abundant and good quality pollen and a low number of ovules per flower can confer relief from pre-zygotic pollen limitation. Endemics are not always disadvantaged: the relative pollination success of endemic and non-endemic species, and its quantity and quality components, are community dependent.
It’s time for the last paper of the pollinator-driven speciation week. The previous posts have examined how pollinators can select and cause diversity among plants but, if they’re all the same species, could pollinators simply mix all the genes back up again?
Floral odour chemistry defines species boundaries and underpins strong reproductive isolation in sexually deceptive orchids by Rod Peakall and Michael R. Whitehead tackles the final part of the process of speciation. Once you have differences between plants, how do barriers to cross-pollination arise? Peakall and Whitehead examined orchids in the genus Chiloglottis, which appeared in the blog last month.
Chiloglottis orchids are found in eastern Australia. The flower doesn’t provide food, instead is appears to offer another reward. Part of the flower looks like a female wasp, and it has the scent to match. Any male attempting to mate will be disappointed not simply because the female is a fraud, but also because the orchid tags the wasp with pollinaria. These get carried to the next destination, which might well be another Chilogottis orchid.
Chilogottis appeals to thynnine wasps, but these aren’t all one species. Each wasp will be looking for specific mate. The way the wasp finds the mate is through the scent. Could variations in the chemistry of the flowers isolate species by attracting one kind of wasp, but not another?
The study initially showed this was plausible. Some of the orchids flowered around the same time of year in the same locations as their sister species, so clearly there was some barrier that wasn’t caused by geography. In addition, there was a lot of morphological overlap between many of the flowers, so there was no mechanical reason why they should be isolated.
It was also possible to cross-breed the flowers when pollinated by hand, and these produced viable seeds so this adds to the puzzle of what the barrier is.
The big observable difference what the chemical cocktails that the plants put out as their floral scent. If you categorise the flowers this way, then there are genetic differences. Peakall and Whitehead argue that what the Chilogottis orchids represent are a number of plants in the process of divergence. This helps give an overview of the divergence process. For example if you visited a forest, you could see trees in all states from saplings to fallen trunks and work out the life cycle of a tree without waiting hundreds of years. In a similar way, being able to identify different diverging branches of orchids at different stages means you get an overview of speciation without having to wait generations for the final result.
To some extent the idea of pollinator-driven speciation could be a puzzle. For many angiosperms, it’s pollinators that keep the species together, exchanging pollen from the flowers of one plant to the other. Peakall and Whitehead’s paper show how the Grant-Stebbins model works, with the shift in pollinators leading to the speciation of the orchids. What might look like a paradox is soluble after all.
Our Grant-Stebbins week continues. Today it’s a question of looking at the geographical context of floral adaptations. Why might you find a plant here and not there? Is it the pollinators that cause it?
Matching floral and pollinator traits through guild convergence and pollinator ecotype formation by Ethan Newman, John Manning and Bruce Anderson examines seventeen members of a pollination guild. These are plants that might look similar and appeal to similar pollinators, but are not closely related. These plants are all pollinated by Prosoeca longipennis a fly with a long proboscis, though the longipennis actually refers to long wings.
How long a P. longipennis proboscis is depends on which flies you observe. It’s found across a wide range of South Africa. The flowers that attract P. longipennis all have similar attributes. They have long corollas, tubes of petals that made the head of the flower. They have similar colours and no scent. Newman et al. also note another feature that makes them specific to P. longipennis – they flower in the autumn.
Observations in the field were able to confirm the existence of a guild and add more members to them. It was then a matter of examining different sites and seeing how they correlated with the local flies. What they found was there was a strong correlation between the tube length of a flower and the average length of the fly’s proboscis at each site. If a site had significantly longer P. longipennis, then the flowers would also have significantly longer tubes.
However, not all of the plants relied on P. longipennis. Some could be found at sites where the flies didn’t visit. Here the same plants would have morphologically different floral displays, showing there was a correlation between the flowers and the local pollinators.
What Newman et al. show with their work is that a variety of plants with different evolutionary histories are all hitting on the same solution. The common factor is they’re all trying to attract P. longipennis to pollinate them. At the same time they also show that while corolla size matters, and colour might matter, the size of the reward doesn’t. Being able to put together a description of the guild means that they can make a prediction. They say that from what they have observed, P. longipennis should visit Watsonia plants, in particular W. galpinii
This kind of prediction reminds me of the Angraecum sesquipedale prediction that Darwin made. em>Angraecum sesquipedale is an orchid that has an amazingly long tube. Darwin predicted that a moth with an equally amazingly long tongue would be found that pollinated it. He didn’t live to see his prediction proved right. Hopefully Newman, Manning and Anderson will have their prediction confirmed a lot sooner. If they’re right they will not just know what fly pollinates the flower. If they know the location they’ll have a good idea of how long its proboscis is too.