Mimicry overestimates and constraints

Mimicry is often the first conclusion jumped to when two coexisting plants resemble each other, or an organism. In many cases alternative explanations have not been investigated, and mimicry should not be established until they have been considered and appropriately excluded (Ruxton and Schaefer, 2011). Alternate hypotheses concerned with mimicry often exhibit fewer restrictions, and would therefore be more likely to allow the plant to evolve. Ruxton and Schaefer (2011) argue that the lack of elimination regarding alternate hypotheses is a massive restriction in current plant mimicry research. If researchers were to more clearly focus on deciphering a mechanism or applying the Ockham’s razor principle mimicry might not always be the hypothesised explanation (Ruxton and Schaefer, 2011). Researchers of plant adaptations need to consider convergent evolution and exploitations of sensory biases as alternative hypotheses to mimicry when attempting to explain similar aspects between plant species (Ruxton and Schaefer, 2011).

Mimicry in plants is also subject to many evolutionary constraints, and it can be concluded that there will always be a limit to a plants mimicking ability. One such constraint is the fact that plants are sessile, which presents multiple limitations (Williamson, 1982). The sessile nature of plants allows pollinators or herbivores time to assess the mimic and the model, and possibly decipher differences and learn how to distinguish between them. Another problem is that the pollinators or herbivores may recall the location that the mimic is in as being undesirable, the organism may have a bad experience and be deceived by the plant and learn not to visit that location again; being sessile the plant cannot change location. Another constraint on plant mimicry is their aggregation habits (Williamson, 1982). Species tend to clump together, a habit which can increase a visitor’s likelihood of visiting too many mimic plants and avoiding the area all together. Aggregation can also occur on the individual plant, with appendages such as fruit, flowers and leaves as the mimic signal transmitters; this aggregation having the same negative effect on a visitor as multiple plants (Williamson, 1982). Some deceptive species, such as the orchids, have evolved to limit aggregation by only occurring in small numbers and over wide spatial distances, so visitors are unable to have negative associations with individual plants. The sessile constraint can sometimes be avoided through the use of floral or seed mimicry, in which the appendages involved are only present for a short period of the plants lifespan. Aggregation constraints are perhaps the largest constraint on plant mimicry because visitor deception is much more successful and efficient if there are not many mimicking plants around, so they can never occur in large numbers (Williamson, 1982).

A variation on leaf variegation

Most cases of plant mimicry involve the mimicry of a single host species, with detailed leaf morphology such as variegation eventuating over many generations. The climbing plant Boquila trifoliolata is different, it is capable of mimicking multiple hosts, even within the same plant. The twining vine is endemic to the temperate rainforest of southern South America, with its leaves variable in size and shape and composed of three pulvinated leaflets that can change their orientation. B trifoliolata mimics the leaves of its supporting trees, capable of copying shape, colour and vein conspicuousness among multiple other features, and is able to change its leaf morphology accordingly when transversing multiple hosts (Gianoli and Carrasco-Urra, 2014).

Variables unsupporting of the mimicking hypothesis of B trifoliolata have been disregarded. Different light patterns can be excluded because the light environment across the vines are generally homogenous, and in all measured cases leaf type did not differ when the light environments were different. Also the leaves of B trifoliolata that were not supported by a host and located on the ground largely differed from the leaves of the plant when it was supported by a host tree, and were similar to the leaves present when growing on a bare tree trunks. It can therefore be established that when there is no leaf to mimic, the supported vines very similar to the unsupported vines, which shows the ‘standard’ leaf phenotype of the species (Gianoli and Carrasco-Urra, 2014).

The mimicry of B. trifoliolata,the vine denoted by a V and the host tree by a T.

 Herbivory avoidance is a hypothesis that can be used to explain why B trifoliolata mimics its hosts. Evidence of this can be seen when comparing rates of herbivory of the host plant, to the mimicking vine; the rates are very similar, especially when compared to the high rates of herbivory observed in the vines when they are not displaying mimicry; unsupported by a host tree. Leaf herbivory was also higher on vines that were climbing unleafed hosts of which they could not mimic, and were displaying their ‘standard’ leaf types. This suggests that the vines are not only avoiding herbivory from climbing a host and by avoiding ground herbivores, but also from mimicking a tree that the herbivores don’t eat. It has even been observed that B trifoliolata can display leaf mimicry without any contact with its host (Gianoli and Carrasco-Urra, 2014).

Herbivory index of B. trifoliolata when on a host tree (a), when unsupported (b), and when on a leafless host tree (c).

 B trifoliolata displays advanced phenotypic plasticity with its ability to resemble several hosts simultaneously, and while there is not currently a mechanism to explain this phenomenon, there are two hypotheses. One hypotheses describes the host plant volatiles in triggering specific phenotypic changes in the nearby B trifoliolata vine leaves, which has been shown as a mechanism in other plant mimicry systems. Volatile compounds have the ability to initiate numerous changes in the plant transcriptome to elicit specific responses in neighboring plants, however this plant to plant signaling has not yet been reported to cause specific morphological changes in leaf morphology and the reprogramming of genes in a mimicking species. An alternative, but less plausible hypothesis involves horizontal gene transfer between plants. B trifoliolata leaf plasticity could involve horizontal gene transfer that may be mediated by airborne microorganisms, which is supported by the fact that mimicry is observed with respect to the vegetation to which the vine is nearest. Further research into B trifoliolata leaf mimicry may lead to the identification of host tree volatiles or even vector mediated gene transfers that can trigger different leaf morphologies (Gianoli and Carrasco-Urra, 2014).

Crop mimicry

The cultivation of crops has imposed a new selective pressure on plants, in particular on weed species. While crops have been genetically modified and selectively bred, many species of weeds have been able to keep up, adapting to the changing morphology and phenology of crop plants to the point where they are almost indistinguishable. This mistaken identity has continued the survival of crop mimics (Vavilovian mimicry); they reap the benefits of fertilisers, pest management and irrigation, all whilst being undetected by the farmer. These mimetic forms of weeds are most likely subject to the pressures of hand weeding practices in Asian crops, and in some cases have been found to be more similar in attributes to the crop than to its closest species relative. This high degree of specialization developed in weed mimicry largely restricts species distribution, as the weeds can only occur where the crops are grown and have often developed to rely heavily on the cultivation conditions. This is considered strange for a weed species which are generally characterised by their ability to live in a variety of environments, natural and disturbed. Some weed populations have even developed the ability to mimic a specific phase of the life history of a crop when weeds are most likely to be removed, preventing eradication (Barrett, 1983).

Mechanization and herbicides are the most influential pressures on crop mimics in the modern era. Chemical selection has vastly altered the composition of crop weed floras, and the increased use and sometimes incorrect use of these chemicals has led to the alteration in the genetic structure of weed populations. There is evidence that weed populations have evolved phonological patterns which aid in the optimization of their survival. For example a weed of maize, Zea Mexicana, has developed a germination inhibitor which prevents the plant from germinating in fallow years, protecting the plant from herbivores (Barrett, 1983).

Crop mimics have developed such strong relationships with their model crops that in many cases they are unable to survive without the crop. As agriculture has become more mechanized and chemically controlled there has been a decline in crop mimics, and it is hypothesised that with the decline of these mimics other agroecotypes of weeds will take their place (Barrett, 1983).

Food or mate ?

South Africa exhibits a high incidence of fly mimicry, with Gazania, Dimorphotheca and pelargonium, species displaying convergent evolution (Johnson and Midgley, 1997). In particular the African daisy Gorteria diffusa, while completely unrelated to other fly mimicking species, depends on the mimicry for pollination. The flowers of G. diffusa actively mimic female Megapalpus capensis flies through olfactory and visual signals. While M. capensis flies are not generally the only visitors to the daisy, they are the main visitor, and are sometimes the only successful pollinator throughout the flowering season (De Jager and Ellis, 2013). G. diffusa are unable to pollinate themselves, so rely exclusively on the success of pollinators.

G. diffusa displaying fly mimicry. Adapted from www.plantsci.cam.ac.uk

The black spots on the daisy are raised and shiny, exuding the scent of female M. capensis flies. Removal of these black spots results in a significant decrease in M. capensis visitations, as does the replacement of these spots with simple ink spots (Johnson and Midgley, 1997). Investigations of these spots using electron microscopy has shown that they consist of a wide variety of complex cell types, and specific epidermal sculpturing may explain the UV reflectance pattern which is remarkably similar to that of M. capensis. This leads to the conclusion that the flies are responding to the incredibly fine detail in the spots. The number of spots per capitulum is varied amongst the species, and can even be varied among the capitula of a single plant, but after analysis it was found that M. capensis does not discriminate between flowers with higher or lower number of spots (Johnson and Midgley, 1997). While male M. capensis are strongly attracted to the flowers because of the presence of the black spots, the females also visit the flowers. 

SEM photographs of G. diffusa ray floret. Adapted from Johnson and Midgley, 1997.

G. diffusa are commonly referred to as sexually deceptive flowers, this is incorrect because they also offer a pollen reward. It is believed that the species has retained its floral reward after evolving fly mimicry because the black spots are used to attract M. capensis to the flower, where it can then feed and pollinate. The pollen reward is hypothesised to be the primary reward with the black spots an added ‘incentive’, which has provided the daisy with an efficient pollination mechanism (Johnson and Midgley, 1997). An alternate hypothesis is that the flies associate the black spots with a floral reward, and that is why those flowers receive a higher incidence of visitation. But this is unlikely, mostly because even young flies are attracted to the spots and there are many instances where the male flies will land on the flower, attempt copulation with the spots, then fly away without feeding on the pollen (Johnson and Midgley, 1997).

'Occupied' mimicry

There are many methods which plants can use to deter herbivores, one way to defend from butterflies is to appear as though the plant is already ‘occupied’. A number Passiflora species have independently developed structures on their leaves resembling the eggs of their main herbivore, the Heliconus butterfly. Their structures are a golden colour closely resembling the colour of butterfly eggs just before hatching, and they tend to cluster at meristems; where the female prefers to oviposit. It is believed that these structures have evolved specifically to mimic the Heliconus butterfly because; they are the primary defoliating herbivore, and the females exhibit great care when assessing a plant for oviposition sites when eggs are already evident. When Williams and Gilbert (1981) tested the validity of this mimicry they found that Heliconus is significantly less likely to oviposit on plants with eggs, regardless of if they were mimic eggs or real eggs (Williams and Gilbert, 1981). 

Although the butterflies did oviposit on some of the egg mimic plants, it was substantially less often than without the mimic eggs. This is because the female Heliconus prefer not to overcrowd a plant to reduce the chance of caterpillar cannibalism if the Passiflora is completely defoliated. Heliconus larva can have a serious detrimental effect on the health of the host plant. In this study, the Passiflora plants subjected to herbivory exhibited suppressed flowering and an increased incidence of root disease (Williams and Gilbert, 1981).

Passiflora exhibiting egg mimicry. (Gilbert, L http://www.mobot.org)

Another slightly different example of what can be categorised as ‘occupation mimicry’ in the butterfly context, is that of the Desmodium motorium plant. This plant exhibits large leaves that perform slow, deep elliptical movements typically at night, and small stipule movements during the day. The leaves partake in elliptical circles every few minutes, and until recently there was no hypothesis as to why the plant had acquired this movement ability. Lev-Yadun (2013) proposes that this movement is the mimicry of a butterfly resting (slow movements) or ovipositing (faster movements) on a leaf. This mimicry, similar to the one mentioned above, is believed to trick passing butterflies into thinking the plant is already being occupied and is unavailable for oviposition (Lev-Yadun, 2013).

This mimicry technique may also be a method of attracting butterfly predators to deter females from landing. The elliptical movements, which are very similar in timing to that of a butterfly, may serve as an attractant to birds, reptiles and arthropods why may subsequently eat Desmodium motorium’s herbivores or serve as a successful deterrent (Lev-Yadun, 2013).

This video shows the (very slow) rotation of a stipule.

What's that smell?

Investigation into carrion and faecal mimicry is not the most popular research topic in evolutionary biology. As mammals, it is ingrained in us to avoid such things because of the potential diseases and pathogenic microbes that may be harbored within them.

It is proposed that carrion and faecal mimicry in flowers has evolved to not only attract pollinators, but also to deter herbivorous mammals (Lev-Yadun and Gutman, 2013). Carrion and faecal mimicry is also hypothesised to give a false warning to herbivores of nearby predators. Flowers mimic the scent of carrion which would likely be surrounded by carnivores, or of predator’s urine or faecal matter to fake the recent passing of a carnivore. This fear of predation has been known to influence the herbivores behavior in such a strong way that vegetation structure can be completely changed (Lev-Yadun and Gutman, 2013). Some hypotheses even go as far as to say that carrion odor may assist plants by attracting large carnivores, which may then defend them from herbivores (Lev-Yadun et al., 2009). Research has shown that the mimicking flowers rely heavily on the volatiles used by flies as cues to locate carrion and faeces (Johnson and Jürgens, 2010). It has been discovered that the oligosulphides emitted by the Mediterranean Arum flower are identical to those emitted by animal carrion and trigger the same electrophysiological response in fly antennae (Johnson and Jürgens, 2010). This shows amazing mimicry of the chemical cues given by carrion to successfully exploit carrion flies.

The primary purpose of the odorous emissions is a hotly debated topic, for example the Rafflesia genus only emits odours during its reproductive season, leading to the conclusion that the attraction of pollinators is the primary function of the chemical emissions (Lev-Yadun et al., 2009). Despite this conclusion scientists argue as to whether the evolution of odours was primarily to repel herbivores, or for the attraction of pollinators. 

Rafflesia spp. adapted from http://liseblog.dk/thailand-2012/

Australia - Deception 'hotspot'

The Australian continent is known for its profound concentration of deceptive organisms. Orchids are the favored deception example in the plant kingdom, with sexual deception resulting in over 11 genera and hundreds of species; complementing Australia’s reputation as the deception ‘hotspot’ of the world.  An interesting paper released late last year by Herberstein et al. (2013), suggests and discusses a few distinct reasons why this might be so.

“Is the prevalence of some deceptive systems a reflection of species diversity?”

This hypothesis questions whether the diversification of deceptive species could be a function of overall species diversity. If this were the case then the overall species diversity of plants would be largely proportional to the diversity of the corresponding deceptive species. Using orchids again as an example, only ~5% of worldwide species occur in Australia, which is a very low species diversity and does not explain the high number of deceptive species. Using orchids and other animal and plant examples, Herberstein et al. (2013) concluded that deceptive systems are not a reflection of species diversity.

“Does deception evolve readily in Australia?”

This hypothesis investigates Australia’s environmental conditions, and its isolation and potential for invasion. Australia is mostly dry, with poor soils and frequent fires, phenomena which have been used to explain the production of nutrient poor biomass (can result in low rates of herbivory) along with ample amounts of sap and nectar (which can lead to pollination by larger animals). These environmental restrictions may have increased the selection for behavioral strategies that minimise the costs to survive, e.g. floral rewardlessness in orchids, as well as fragmentation of species due to fire regimes promoting selection for gene flow across larger distances. If harsh environmental conditions are a driver for the evolution of deception, that is where a high number of deceptive systems are likely to occur. Due to limited information in the literature, Herberstein et al. (2013) urge further research into the frequency of deception in harsh environments to help support or disprove this hypothesis.

Australia has been subject to a long history of evolutionary isolation and recent invasions, which may explain a radiation of deception. If this were the case, then it can be predicted that most deceptive species have only recently arrived in Australia, and seek to exploit naïve and endemic species. Australia’s Mediterranean and tropical climates have given rise to high levels of endemism and biodiversity; which provides a broad variety of potential deception targets. Isolated populations are typically vulnerable to exploitation by invasive species, the mixing of invasive species with existing species can lead to new symbiotic relationships; and perhaps even deception relationships.

“Does Australia’s intellectual and research culture encourage discovery of deception?”

Herberstein et al. (2013) discuss their final hypothesis of high levels of Australian deceptive species as being the result of recent research popularity. They predict that the popularity of certain research areas leads to a higher than expected reporting of similar phenomena; including the discovery and investigation of new deceptive species. For example, behavioral ecology in Australia is thriving, especially when the number of research institutions are compared with other nations. Over the years 2010 and 2011 8% of papers published in Behavioral Ecology were authored by Australian researchers and institutions, compared with 16.5% of papers being from the United Kingdom, which has 3 times the number of institutions as Australia. However, Herberstein et al, (2013) issue caution when presenting this hypothesis. This is because a more thorough investigation and analysis on the number of publications on deceptive and non-deceptive systems is needed for a comprehensive evaluation of the high number of Australian deceptive species due to potential publication bias.

Descriptives of the higher education environments in Australia, Europe, United Kingdom, United States of America, Canada, and Japan, indicating Australia’s surprisingly large proportion of behavioral ecologists and consequent contribution to the journal Behavioral Ecology given the relatively few research institutes. Adapted from Herberstein et al. (2013).