Well if this doesn’t beat all! There are flowers that mimic insects, and insects that mimic flowers, and even plants that mimic stones (Lithops) to hide them from predators, but this is the first time I’ve heard of a plant mimicking another plant. Not only that, but the mimic, a vine, can modify its leaves to resemble those of at least eight other trees on which it climbs.
The article describing this, in press in Current Biology (full reference and link to abstract below; the paper is behind a paywall though judicious inquiry might yield a copy), shows that the neotropical vine, Boquila trifoliata from South America, can not only mimic at least eight different species of tree, but—and this is truly amazing—a single plant can mimic the leaves of several different trees if it happens to entwine around more than one of them. In other words, each plant has the genetic ability to somehow sense which tree it’s on, and modify its leaf shape to resemble the leaves of its “host.” Now that’s what I call plasticity!
Below is a screenshot of figure 1 from the paper showing the vine’s leaves (arrows) next to the plant on which it climbs or is near. I’ve added the caption for those with more serious botanical interests. Note how closely the vine leaves resemble the tree leaves. The plasticity involves changes in shape, size, and color.
The other amazing thing is that the vine doesn’t actually have to touch or climb the tree whose leaves it mimics; it only has to be near it. That means that the vine has to somehow sense what tree is nearby. That rules out explanations for the mimicry involving physical contact. But more on that below.
1. What’s the advantage to the vine of being able to modify its leaves to match those of its host tree? The first thing that comes to mind is protection from leaf-eating insects. This could occur in either of two ways, though the authors don’t mention these alternatives.
The first advantage comes if the leaves of its host are somehow toxic to herbivores, who then learn to avoid them. In this case the mimicking vine would be a Batesian mimic, an edible species that takes advantage of a learned avoidance response by the herbivore. (The herbivore could use visual cues, which must be the case here because of the visual resemblance of leaves, as well as other olfactory or other chemical cues, which weren’t investigated in this case. Could the vines show “chemical mimicry” as well?
The other hypothesis is simply that by mingling your leaves with those of an edible, and resembling them, the chance of you being nommed by a herbivore is lessened: most likely the herbivore will go first for the more numerous leaves of the tree, so the vine gets protection by being outnumbered.
The authors, as I said, don’t distinguish between these theories, but they did preliminary experiments to show that the mimicry does seem to confer protection against herbivores. They did this by looking at how often the vine’s leaves were munched it was attached to a tree whose leaves it mimicked, compared to vines that were either naked on the ground or entwined around a tree that was leafless. After showing that the rate of herbivory on vines climbing leafy trees was similar to that of the trees themselves, they also showed that the rate of herbivory of vines on the ground or naked trees was significantly higher.
Now that doesn’t show that the mimicy itself confers protection—only that being on a leafy tree confers protection. The authors still need to show that making your leaves resemble those of the specific tree confers greater protection than if your leaves are mismatched. That could be done fairly easily, I think, through transplant studies.
2. How the hell does the vine know how to grow its leaves to mimic the nearest tree? This is the real stunner, for each vine apparently “knows” how to change into the best of at least 8 possible leaf shapes; that is, the vine carries genetic information to sense the leave morphology of the nearest tree, and also the genetic information to transform its leaves into that particular shape. The evolutionary scenario for how this could happen boggles the mind, for it involves cues and switches between at least eight discrete morphologies. And the mechanistic basis is unknown. How do the vines sense what tree is near?
The authors offer two hypotheses, one much better than the other.
a. Volatile compounds emitted by the tree are sensed by the vine, which uses that signal to change the shape of its leaves appropriately. This is feasible becasue such volatiles are known in some plants, and are used to deter herbivores or affect the grown of nearby plants of the same species.
b. The second hypothesis is far more speculative (and to my mind, unlikely). Here it is in the authors’ words:
“An alternative hypothesis, but perhaps less plausible, would consider horizontal gene transfer between plants, a phenomenon that is increasingly being reported. These cases include both single and multiple transfer events per species, which are hypothesized to be mediated by a vector or result from plant-plant parasitism or natural grafts. The plasticity in leaf mimicry in B. trifoliolata could involve horizontal gene transfer on an ecological timescale and might be mediated by airborne microorganisms. The latter speculation is based on the fact that mimicry is observed with respect to the foliage to which the vine is nearest, irrespective of whether this foliage belongs to the host tree that the vine has climbed. Further research on leaf mimicry by B. trifoliolata might lead to the identification of the host tree volatiles or vector-mediated gene transfers that trigger differential gene expression in this singular climbing plant.”
I think this far less likely, because I find it implausible that the very genes carried by some kind of vector microorganism would include those involved in leaf shape, and would insert in the appropriate place in the vine genome and be expressed properly. Were I the investigators, I’d concentrate on the first hypothesis.
Clearly there’s a lot more work to be done on this system. But it’s really a fantastic one. Who would have thought that a vine could act like a chameleon, able to change its leaf shape to match the surroundings, to match at least eight different model hosts, and to match more than one host on a single vine? This is the kind of stuff that gave rise to the adage in my field, “Evolution is cleverer than you are.”
___________
Gianoli, E. and F. Carrasco-Urra. 2014. Leaf mimicry in a climbing plant protects against herbivory. Current Biology, in press. http://dx.doi.org/10.1016/j.cub.2014.03.010
Woah…mind asplosion.
…incidentally, and totally off-topic, “Plastic Fantastic” is the affectionate nickname of Canon’s cheapest lens, which is what I first thought of when I read the headline….
b&
And of course Plastic Fantastic Lover was one of the tracks on Surrelistic Pillow by Jefferson Airplane
Indeed.
b&
These vines are … creepy.
Groooan.
Waiting for Chopra & company to offer that the plant changes its appearance by thinking about altering its DNA.
Yeah, but I bet Hart would still say the plant isn’t smart enough to spot the nightjar.
b&
This ‘salad’ does not seem to be stupid at all!
But that’s just it: Hart has already demonstrated himself utterly incapable of gauging intelligence….
b&
Mimicry of any stripe in the natural world is probably the one thing that amazes me the most, and this story might just take the cake. Crazy.
This is one of the most mind-boggling stories I’ve seen lately. If it is real, my guess is it evolved because there are probably specialist herbivorous butterfly larvae that eat the vine, like the Heliconius butterflies whose larvae eat only passionflower vines, and so there is an advantage to polymorphism and disguise. In Passionflower vines, different species have evolved dramatically weird leaf shapes to disguise them from female butterflies visually searching for Passiflora vines in the forest. (They have even evolved butterfly egg mimicry!)
But I have to say I am a bit skeptical of this….there are major opportunities for observer selection bias effects here. I haven’t read the paper, though, and I suppose the author checked for that. There is also the potential for the environment (sunny or shady, windy or calm) to determine which trees grow in a particular spot, and the vine could independently be adjusting its leaves to those same environmental variables (many plants do that). The tree species that outcompetes its neighbors in a given environment will also tend to be one whose leaf structure is optimized for that environment, so if the vine has plasticity in response to its environment, its leaf shapes may correlate with the local trees, even though there is no causal influence from tree leaf to vine leaf. Many vine species are very plastic with respect to their environment, because they generally start very low in shady, humid conditions and end up in bright sunny habitats when old. Again, Passionflowers are good examples; their forest understory leaves often look nothing like the leaves of the same plant when it reaches the light.
I should add that plants do mimic each other often in the tropics. For example, there are Elleanthus orchids that mimic bamboo in my yard here in Ecuador, and shrubby Melastomes that mimic begonias (or vice versa, or both). But these are hard-wired effects, not plastic.
Even if they are sensing microclimates rather than proximity to specific trees, there could still be selection pressure for mimicry of trees typically found in those microclimates.
Sure, there could be, but my point is there doesn’t have to be. And even if there were, there is no need for the vine to detect what kind of leaves are around it.
The claimed detection of surrounding leaf shapes or types is the tricky and surprising part of this story, and it is very hard to think of a plausible mechanism for this. I think the described effect can be achieved with no detection mechanism at all.
I don’t think anybody’s claiming that the vine somehow sees the shapes of the surrounding leaves and actively mimics those shapes the way a chameleon mimics its background.
But clearly the vine has some mechanism for detecting different environments or microclimates and generating the appropriate leaf shapes. Natural selection then optimizes leaf shape for each environment, and mimicry could plausibly play a role in such optimization.
Seems to me that (for instance) the spiny leaf tip in photo H would be hard to explain as mere convergence rather than as mimicry.
What people are claiming is that the vine recognizes (through chemistry, or whatever) the types of plants that surround it, and mimics them. I still think that is less likely than the alternate explanation of shared environmental causes (plus lots of plasticity in the vine). Correlation, not causation.
Spiny leaf tips and other spiny plant parts are fairly common in plants of drier habitats, and in those communities it is most likely a result of convergence rather than mimicry.
But surely “environmental causes” include the proximity of other plants, the shadows they cast, the chemicals they emit, and so on. So I guess I’m not seeing a sharp line between environmental cues and direct detection of other plants.
Yes, that was ambiguous. I’m trying to distinguish between general non-specific physical environmental cues versus species-specific direct cues about the kinds of plants that are in the neighborhood. I think it is possible to have mimicry without the vine being able to distinguish the shadows the host casts, or chemicals or genes it emits, or mycorrhiza it shares, or any other causal signal from host to vine. Merely by responding adaptively to the local physical environmental variables, the vine could converge on the same leaf shape that its host uses, because the host that survives in a given spot also most likely has leaf shapes optimized for that local environment.
I too have not read the paper, but I found the supplementary material and am willing to accept for the time being the referees judgement at Current Biology.
The vines were studied in temperate Chilean rainforest. These forests have few tree species; 12 were listed by the authors. A genetic tendency for a vine to become mimetic through induction would be selected more strongly in a species-poor forest than in one rich in tree species. With few tree species, generations of vines will commonly find themselves growing on one of a small number of leaf-environments, and mimetic selection would be highly directionally. The temporal sequence of evolution of mimetic leaf shapes may correspond to the abundance of the trees that are mimicked. The first trees to be mimicked were likely common species (large selective coefficients) and rare species added. The accuracy of mimicry might be expected to be better for tree species more commonly used for support. There are lots of interesting hypotheses.
Authors Fernando Carrasco-Urra and Ernesto Gianoli-Molla are to be congratulated for their perceptive botanical insights. Ecologists in temperate northern clines should be on the lookout for convergent situations.
I mostly agreed with that, but have doubts on one point: “If it is real, my guess is it evolved because there are probably specialist herbivorous butterfly larvae that eat the vine, like the Heliconius butterflies whose larvae eat only passionflower vines, and so there is an advantage to polymorphism and disguise.”
Unlike Passiflora, this is apparently a very rare vine and I wonder if it has specialist herbivores. Many rare things don’t — the “island” is too small.
Several years ago (c. 2008) I received material of Boquila in a large set of specimens from Chile, but it was unidentified. I didn’t know what it was so sent a sheet to MO Bot. Gdn. for identification — they IDed it, and sent me a nice note saying it was the first material of this genus they’d ever received. They had no comparative material and IDed it from the literature. The sheet I sent them seems to still be the only specimen they have, according to their website db. MO, for any who don’t know, is one of the largest herbaria in the world and very active with the flora of Latin America. If Boquila was at all common, I don’t see how MO could not have had it previously.
That’s interesting. The paper’s authors seemed to have found lots of examples, though. I of course have no idea if there really is a butterfly that specializes on this vine. But that kind of herbivore could drive the mimicry if it exists.
In the supplementary material, the authors identify weevils and leaf beetles as important herbivores. Leaf beetles are candidate selective agents.
OT–this reminds me of an analogy of Mark Moffett’s, from a talk he gave years ago, to the effect that vines are like slow snakes; they move to where they can bask in the sun, and if the canopy changes so as to shade them, they just move on to the next sunny patch.
Not having seen the paper, I would like to see a large sample of B. trifoliata leaves plotted in a shape space to see how discrete the morphs are.
Seen the paper now; they didn’t do that. Can I petulantly call it a ‘failure of peer review’ now?
What I meant is, if there’s just correlation within a fuzzy continuous distribution of shapes that would be quite impressive, but one-to-one matching of discrete morphs to hosts would be another thing.
I have to agree with this assessment–I am not completely blown away by the degree of mimicry. The resemblance seems mostly about leaf size, not complex shape measures. Is is possible that local light and microclimate conditions similarly affect these leaf parameters in both the host and vine plants? If so, this would not be mimicry.
Mostly about size? Did you look at the picture? Size, shape, color, and even texture are all duplicated.
Are you really suggesting that the samples in the photograph are cherry-picked, and just coincidentally show an incredible match between vine and host?
Of course the photos in the article would show the best examples, so naturally they’re cherry-picked! That’s the opposite of coincidence.
Table 1 in the paper lists correlation coefficients r^2 for 11 leaf measures between vine and host, of which six are considered significant, but five of these (maximum length, width, area, perimeter and petiole length) are just measures of size and obviously not independent. Colour shows no significant correlation, and texture was not reported, so you’re claiming more than the authors of the paper.
Testing vines on leafy trees vs. bare trees seems like a poor experiment to me.
Could the vines be detecting what tree is nearby through the roots?
That wouldn’t explain the occurrence of vines whose leaves match multiple hosts, would it?
I like that idea. Few realize that there’s generally as much plant below the soil as above, and that most of agriculture is keeping the roots happy.
It’d be a pretty straightforward lab experiment to test it. Put the plants in pots, and train the vines onto hosts in other pots.
b&
Very interesting, but I do concur with comments above that the range in plasticity is in the range we see in plants due to local environmental conditions. Poison ivy is my favorite example of this, making it hard to identify that plant.
If there is indeed mimicry going on here, I propose a hypothesis c: that both the form of the host trees and the vines are influenced by mycorrhizae. These are the soil fungi that form close, symbiotic associations with plants by attaching intimately with the roots. Some are pathogenic, but others provide beneficial extensions to plant root hairs & they can even supply their plant partner with nutrients. It well known that mycorrhizae influence the form of the aerial parts of plants.
The host trees and vines could share a common mycorrhizal species and clone, and this could influence their above-ground form in similar ways so they look similar. I just made this hypothesis up, but it is a lot more plausible than hypothesis b.
The impression I get from the excerpts is that the trees being imitated ALWAYS look that way; species-specific characters that don’t depend on presence or absence of mycorrhiza.
As I wrote under Comment 4, there doesn’t have to be any causal influence at all from the tree to the vine in order to explain this, unless there is more to the story than I’ve seen in the short excerpts from the paper.
Agreed on possibly no causal influence, thanks for pointing that out, it is simpler.
I am not at all attached to my hypothesis — I am just musing — but the form of the host tree leaves may not in fact be entirely species specific. The forms of plants are very often within a range called a ‘norm of reaction’, where genotype and environment (in this case, mycorrhizae) can combine to give plants a particular form. The host trees could have a particular, species-specific form b/c they are restricted to normally grow in a certain environment, and encounter certain mycorrhizae there. One species may grow near river banks, others in areas of higher elevation, and so on. The vines may be less restricted in the environments they grow in.
I was thinking mycorrhiza too, but as long as we are speculating from known biological entities I would add vesicle communication as hypothesis d.
It was recently found to happen between cyanobacteria in addition to earlier bacteria, in the form of “continuously release lipid vesicles containing proteins, DNA, and RNA.” It is believed to help with nutrient storage/sharing, defenses against viruses et cetera, “facilitating interactions between cells and their environment from a distance.” [ http://www.sciencemag.org/content/343/6167/183.abstract ]
Is there any reason not to believe chloroplasts, perhaps mitochondria or even the eukaryote cell/multicellulars do any of this? The trait could have been appropriated any time.
You’re overlooking the part where “a single plant can mimic the leaves of several different trees if it happens to entwine around more than one of them.” I don’t see how any hypothesis about roots or soil conditions can explain that. So we’re pretty much limited to hypotheses about the immediate environment of the leaves.
Or the immediately localized mycorrhizae. Just say’in.
No, that apparently can’t be, because it seems that a single plant can grow through the crowns of several trees and the leaves will change form from one “host” to the next. Perhaps these are separate branches entering the various trees, but apparently they’re all parts of one individual. The plant (vine) is still presumably rooted in just one place and so its roots are exposed to the one set of mycorrhizae that exists in that vicinity. If things are as reported.
But, I find this report very puzzling. Things may not be as reported. I wonder if important information is unknown or if something is wrong somewhere in the description of the vine’s behavior.
That is, I agree with Gregory. “So we’re pretty much limited to hypotheses about the immediate environment of the leaves.”
And I’d add that it’s the immediate environment of the leaves at the time of their expansion or initiation that should matter. Mature leaves are not going to change shape even if you disentangle a vine shoot from one tree and drape it on another. At least, I’d be shocked if they reconfigured themselves.
The bit about growing through different levels of a canopy, taking on a different shape at different vertical levels of growth, of course would kill my earlier idea. I was thinking of a more ‘horizontal’ arrangement where the vine could be rooted alongside more than one tree.
@Torbjörn: Off-topic here, but wondering if you saw my query about your “Greenpeace are terrorists” claim here? (I never tick the ‘notify’ box as it hasn’t seemed to work before, but here’s trying again)
“a single plant can mimic the leaves of several different trees if it happens to entwine around more than one of them.”
This would seem to eliminate the environmental and root hypotheses, I think.
Is it possible that it just shoots out all 8 kinds of leaves, and whichever ones don’t get eaten flourish? Or that it tries variations if its leaves are getting eaten?
I haven’t seen the paper, but this find is just begging for some controlled experiments.
This doesn’t eliminate the environmental hypotheses. A single individual vine can (and usually does) span several different environments as it makes its way upwards.
In the supplementary material the authors emphasize the uniform shadiness of the forest. Leaves were sampled from stems near the ground, height 80-160 cm. It seems unlikely that some physical cue is so tightly associated with tree species that it could generate a strong statistical association. Although the authors do not say it, it appears that vines growing up tree trunks, from deep shade to full sun, vary little in leaf shape.
Hmm, that’s strong evidence against my hypothesis….
I have noticed that the deer around here seem to consider vines an attractive food item.
So perhaps beyond merely safety in numbers, the vine is getting safety in obscurity – the herbivore would target the vine if it recognized it, but would rather keep looking than go for the edible (but relatively unappetizing) tree.
As for the idea being floated that similar environmental conditions drive the similarity in leaf shape, I’m seriously skeptical. Just how much difference in leaf appearance can there be under different environmental conditions? Why do different plants in the same environment have dramatically different leaves? I’m no botanist, but my own experience with local wild plants and grown cultivated plants just doesn’t jive with the similar environments theory.
There can be a huge amount of variation in leaf shape based on things like day length at the time of expansion. Ipomoea is famous for this, for example. Sun/shade has a big influence on blade width and thickness in many species. On pubescence density too.
But, the correlation of the sort of leaf shape details to particular hosts is hard to explain that way, IMO.
There are a number of examples of convergence of unrelated plants within a particular region and habitat.
Best that come to mind are brownish, divaricate-branched, small-leaved low-growing shrubs. Almost certainly converging to make it hard for moas to strip leaves. ?Brownish — who knows, but it’s very common there.
Another anti-browsing mode is “wire-netting”, protecting new growth within a cage of branched, interlaced woody structures in low-growing Mediterranean plants — usually usually formed by lignification of persistent, dead flower stems from the previous year. [But sometimes other plant structures are involved — but the pattern is quite distinctive and to my knowledge it’s uncommon in other floras] Undoubtedly most effective against precision browsers like goats. Again, this involves many unrelated plants — I’ve seen it in as many as 10 different plant families on a single hillside in Crete.
It seems a little odd to read that butterflies and other invertebrate predators of this vine are orienting visually — usually chemical cues are dominant. Vertebrate browsers, present or extinct might play a role, perhaps??
I made that comment about butterflies. It’s well-known that some butterflies use visual cues from a distance to find their hosts, though once they find them, they tap them with their legs to sense their chemistry. As I said, this is an important driving force in Passiflora vine evolution, driven by the very visual Heliconius butterflies.
As for environmental variability, there are many examples of how leaf forms change dramatically depending on light levels. Some vining aroids are excellent examples; the shade-grown juvenile leaves are undivided and pressed tight to the host trunk, but once the leaves get into the light of the canopy they are deeply divided and not appressed.
Amazing.
I have no idea what the advantage is in this case, but there must be one to mimicking the host’s leaves because there are several other cses like this (though none as impressive). Australian mistletoes, for example, mimick the leaves or twigs of their hosts, be they Acacias, Eucalypts or Allocasuarinas.
Having carefully thought through the alternaitve explanations posed here, I am still not exactly convinced. I can imagine some correlated response to microclimate between host and vine would cause some patterns, but they seem too strong to me. Mycorrhizae seem unlikely because specialization is often low for AMF. For the vine specialization might be beneficial, but for the host less so.
Could it be that we are actually looking at different cryptic species/ecotypes? Then each of these has 2/3 different morphs, depending on host recognition/environmental cues, of which the “non-host” state is so similar that we confuse them to all belong to the same species/ecotype?
Reblogged this on Mark Solock Blog.