Centaurea
(Knapweeds)
Centaurea debeauxii (Chalk Knapweed) as seen in the British Isles (southest England). This form with strongly recurved phyllaries (scale-leaves around the capitulum) was termed var. microptilon in the older literature, but the degree of curvature likely depends on humidity as we shall see.
This is a complex topic and at the moment this article is a work in progress. This is part of an ongoing project to study Centaurea.
The Centaurea or Knapweed genus of Asteraceae displays some very interesting biology and it is a truly fascinating genus.
Click on images for full-size view.
Above: the Centaurea debeauxii x C. nigra form eradiata normalis).
Centaurea is a large genus of several hundred species, though the exact number depends on classification as Centaurea tends to form hybridizing complexes of several species. It is found in the Northern Hemisphere. We will begin by focusing one one species complex: the C.nigra aggregate, which in the British isles consists of Centaurea nigra (Black Knapweed), Centaurea debeauxii (Chalk Knapweed) and Centaurea jacea (Brown Knapweed). The latter, however, is an occasional import from the European continent and was once more widespread in the British Isles, but has largely disappeared by breeding into the other two species, though not entirely.
Above and below: the same capitulum. This one is perhaps close enough to C. nigra to be classed as such but is more likely C. debeauxii x C. nigra when all things are considered. One problem that confronts the study of Centuarea is that some of the designated species are thought to be hybrids by others and hence there is no consistent taxonomy or naming of the species. Furthermore, some forms designated as species by some authors, are designated as varieties or subspecies by others.
Additionally, descriptions given of continental European types, as given for example in Flora Europea (irrespective of type specimens) have significant morphological differences to populations in the British Isles (e.g. in numbers of pectinations on the phyllaries) with the same species designation. I suspect that it may be more fruitful to define species/subspecies/varieties by considering geographical region rather than trying to define continenetal-wide taxa. Some support for this idea comes from a study by Hilpold et al. (2014) in which it was found that geography predicted phylogeny better than morphology on molecular grounds.
Note that that the flower head or capitulum (plural capitula or capituli) bears many florets as is typical of the compound or composite flowers of Asteraceae where several component flowers (florets) usually combine to form a flower-like capitulum (pseudoflower). In this example, almost all of the florets are tubular disc florets with very few peripheral rays (some appear to occur at the bottom) hence 'eradiata' as in lacking rays; though rays (pseudorays) do occur in some forms. Such a rayless capitulum, without modified peripheral florets, is also referred to as discoid.
Rays are peripheral florets modified for display (and generally neuter or sterile) with elongated petals, radiating outwards like the corona of the Sun. Each floret has a petal tube (corolla) made up of 5 petals fused for much of their length but usually separating into 5 distinct lobes at their tips. In ray florets a number of these lobes (often 3) unite into a strap-shaped ray (or ligule) - the remaining petal-lobes being reduced - recall that a 'ray' is a straight-line, so these ligules radiate outwards like false petals around the compound flower like the rays of a star. In Centaurea, however, when rays do occur they are generally mere extensions of the flower tube and some of the lobes which do not unite into a ray-like ligule but simply elongate and splay apart and hence are often termed pseudorays.
Above and below: an individual with clear rays (form radiata).
Marsden-Jones & Turrill (M-J & T) in their classic work 'British
knapweeds' suggest that this may indicate the presence of genes from Centaurea
jacea. Many years ago southern England saw an influx of Centaurea
jacea from the continent, as a grain-seed contaminant, and this
hybridized with both the C. debeauxii and the C. nigra.
However, with removal of these seed contaminants. C. jacea in
Britain has declined but its genes persist in some Centaurea
populations.
There are different forms of Centaurea nigra, C. jacea and C. debeauxii, some of which depend on the characteristics of the florets. In radiate forms (form radiata) the outermost or peripheral florets are modified into rays (or pseudorays) which helps advertise the capitulum by making it more striking to catch the attention of passing pollinators. These outer florets are, however, generally neuter and so are unable to set seed but have become specialized only in display.
There is a more detailed classification based on absence or present of rays (pseudorays): eradiate forms have none, radiate forms have peripheral florets with ray-petals 2.5 cm or more in length, whilst the semiradiate form has rays less than 2.5 cm in length. I did not measure this specimen, but it looks radiate to me. Further complications arise due to variations in the length and gender of the florets, as discussed below.
Above: a rayed form of Centaurea debeauxii x C. nigra.
Above: form eradiata, lacking rays totally. Though close to C. nigra this is likely C. debeauxii x C. nigra. Hybrid names generally include individuals showing clear signs of introgression. Introgression is a process in which a hybrid crosses with one or both parental species over a number of generations, producing a form close to one of the parents. There comes a point where the classifier may place such a specimen in either parental taxon, if it is deemed sufficiently pure but where the boundaries are drawn is subjective. Perhaps clear morphological boundaries should be defined for the purposes of mapping species distributions?
Pollination Mechanism
The florets of many Centaurea (and indeed many asteraceae) show interesting movements in response to the touch of a potential pollinating insect.
The following chiefly describes the mechanism in Centaurea americana (a native of Texas) but the mechanism is similar in the Centaurea nigra group. The florets produce nectar to attract pollinating insects. When each floret (excluding ray florets when they occur) opens the pair of stigmas are pressed together, concealing their pollen receptive surfaces and the style is enclosed by the ring of fused anthers. The anthers shed their pollen into the common cavity enclosing the style. The style has a ring of hairs around its top, just beneath the closed stigmas. These hairs are angled upwards. When ripe, the valves at the end of the anther tube open, exposing a few pollen grains, as a further incentive to pollinators ( many of whom eat some of the pollen grains).
A light touch of the anther tube causes the stamen filaments to contract, forcing the style out of the anther tube like a piston, enabling the ring of hairs to sweep up pollen and present it to a visiting insect. You can see bunches of white pollen grains being presented in the picture above. If you lightly stroke a floret which has not yet presented pollen but is ready to then you will see a small mass of white pollen emerge from the end as the folded stigmas protrude. Thus, the ring of hairs acts as a pollen brush. Furthermore, if the female pistil (style and stigmas) is then touched, it will circle from side-to-side or make a circular motion, at least in Centaurea americana, 'searching' for a potential insect to brush pollen on to. Once the pollen is removed the responsiveness ceases: touching an exhausted floret with the stigmas protruding but no pollen left to present will induce no response.
Above and below: the same capitulum. Note that the central florets have not yet opened. This specimen could pass for C. debeauxii but possibly has some C. nigra genes in it.
Above and below: Centaurea debeauxii x Centaurea nigra. These views are intended to highlight the florets and their pollen presentation. 9However, the phyllaries can be easily seen on the full-size image despite being in shade. Note that n this specimen the phyllaries are covered with white arachnoid hairs.
Above: click on this image to view full-size and notice the unfused bases of the stamen tubes which are clear due to the short petal tubes in this specimen, which is a short-flowered form (form breviflora). Below: the same individual. (Photographed on 6 July).
Note the presentation of some of the pollen to a potential pollinator at the tips of some of the anther tubes. Once the pollen is exhausted, the style grows upwards and the stigmas separate, arching outwards to expose their receptive surfaces which are roughened by minute papillae to catch any pollen subsequently delivered to the flower. The flower has now switched from a male stage to a female stage, helping to reduce the likelihood of self-pollination and promoting cross-pollination which is good for the gene pool.
Above: Centaurea debeauxii x C. nigra showing pollen presentation.
The Taxonomic Problem of
the Centaurea nigra Group
Centaurea is a notoriously complex and hard genus to study because it is both considerably diverse but species boundaries are also rather open, with species frequently hybridizing. Distinguishing a hybrid between two parental species is often hard enough, but when three or more species may have contributed to a population's genetics it gets even tougher. However, Centaurea tends to fall into groups, where crossings between groups are rare or non-existent. This means that we can focus, for example, on the C. nigra group (or perhaps equivalently the C. jacea group, also these may also cross with the Centaurea phrygia group as these two groups appear closely related). This complexity is thought to indicate that this genus is recently evolving and that the species may separate more fully over time. The greatest diversity in Centaurea is seen around the Mediterranean, where this genus likely began its diversification.
Centaurea nigra generally flowers early (Jun to early July) whereas Centaurea debeauxii generally flowers late (late July to September) however there can be some overlap allowing the two species to hybridize. The time of flowering is of course affected by weather and latitude (Centaurea nigra flowering later further North) but is also under genetic control with some hybrids flowering at intermediate times in the size, such that a mixed population may flower throughout the summer.
Above: this specimen is close enough to Centaurea debeauxii.
Note the scale-like bracts or phyllaries that form the enclosing cup or involucre (also called the pericline) and notice how the appendages at their tips recurve or arch backwards in this specimen. This happens in dry weather, in damp weather the phyllaries close. When open the florets are also open, making their nectar available to pollinators. In fruit, the fruit are readily dispersed when the phyllaries are open in fine weather. Such movements in a plant part which are passive and occur in response to moisture content are called hygroscopic movements and have many functions in the plant world.
Above: Centaurea debeauxii x C. nigra (Black Knapweed). The phyllaries are rather narrow and triangular for C. nigra and this plant is quite late flowering (end of July) so it likely has some Centaurea debeauxii genes in it, indeed it is from a mixed population. However, using Harris' (2017) phyllary overlap scale this specimen is close to C. nigra. It could probably be classed as either the hybrid or C. nigra depending which characters are given priority. (Similar specimens can be found in the BM herbarium). Interestingly, the hybrids used in Harris' 2017 study predominantly had the phyllary shape of C. nigra. It is not known how these genes segregate - hybrids are not always morphologically intermediate in all characters. The specimens shown here are from Kent where both species occur and hybridize freely and given reports that C. debeauxii is likely to be spreading I doubt that any of the specimens presented here are pure C. nigra and so could be classed as hybrids.
The phyllaries open and spread when dry, allowing the florets to open to pollinators and any ripened fruit to be released. this is easily demonstrable on a specimen like the one above, after the capitulum has been picked. When dampened the phyllaries close and the pericline become tight and compact. Note that in Centaurea nigra the dark tips (appendages) of the phyallaries largely conceal the green or whitish basal parts of the phyllaries.
Dispersal in Knapweeds. For a pdf essay on dispersal in Centaurea, with a discussion of mechanisms, ploidy-levels, ecology and invasiveness see: Dispersal in Centaurea.
The degree of arching outwards of the phyllaries (strictly their terminal appendages) was once used as a taxonomic character, and perhaps there is some real difference in responsiveness or the extent of hygroscopic movements in different taxa, but this would have to take account of meteorological conditions to be comparable across taxa. A decidua form has also been recorded, in which the phyllary appendages are naturally shed (in whole or part).
The form of the phyllary appendages is the chief key to species identification
Above: Centaurea debeauxii.
In Centaurea debeauxii the phyllary appendages tend to be narrower but the appendages also provide a sparser covering to the involucre, with the green phyllaries showing through readily, especially in the lower third of the involucre. In contrast, in Centaurea nigra the pectinations or fimbriae of the appendages overlap strongly and the appendages form an almost complete covering to the involucre.
Above: the pericline (cup-like base of the capitulum) is covered in bracts or scale-like leaves called phyllaries. Each phyllary is divided into the basal green part (often simply called the 'phyllary') which is sometimes yellow or red or covered in white arachnose (cobweb-like) hairs, and the terminal appendage which is usually brown or black, with a paler margin.In this specimen the phyllary pectinations overlap but did not generally extend far enough to cover the central appendage discs so this is likely C. debeauxii x C. nigra but could be considered close enough to Centaurea nigra to be classified as such..
Above: Centaurea nigra. (Flowering on the 24 June).
The appendages have either more-or-less entire or divided margins, depending on species. The phyllaries near the top (the innermost phyllaries) tend to have more entire margins, lower down in the middle rows the margins maybe irregular (as in Centaurea jacea) or divided into a fringe of fimbriae or more regularly divided into comblike pectinations (as in Centaurea nigra and Centaurea debeauxii).
Above: Centaurea nigra. (Flowering on the 24 June).
Thus, there when describing the morphology of the appendages there is a spectrum from entire-irregular-laciniate-fimbriate-pectinate. Laciniate means that the margins are cut up more coarsely than fimbriate and into irregular pointed lobes (giving a torn appearance). Those in the specimen above are clearly pectinate.
In modern usage the term 'pectination' is used interchangeably with 'fimbria' (plural: fimbriae) as the distinction is not absolute but rather subjective. Fimbriae and pectinations may also be called cilia and the margins are then said to be ciliate.
It's all in the degree of fimbria overalp in the middle rows of appendages
In the specimen above, the pectinations of the middle rows of phyllaries overlap with the central discs of the neighboring appendages which is diagnostic of Centaurea nigra (Black Knapweed). Typically one examines the 3rd and 4th rows of phyllaries from the bottom (which is not always obvious since the phyllaries are arranged in a spiral rather than in neat rows). At least this is the approach developed by Harris (2017) and adopting a similar approach with the specimens in the British Museum (British & Irish Herbarium in the Natural History Museum, London) this seemed to be the most robust method.
The specimen below shows a quite different pattern with narrow appendages with short fimbriae:
Above: the middle rows of phyllaries have their fimbriae (pectinations) barely touching, leaving much of the green phyllary bases exposed, diagnostic of Centaurea debauxii. In British populations, another feature to consider is flowering time: C. nigra typically begins flowering in June whilst C. debeauxii generally begins two months later, with a short period of overlap between the two. Caveat: flowering times of plants in general can vary considerably from year to year and with latitude so flowering times given in standard texts are only approximate.
Above: this plant has characteristics suggesting it is a hybrid (or descendant thereof) between C.nigra and C. debeauxii. The fimbriae of the middle appendages overlap to some extent but not enough to cover the green phyllaries. The capitula of Centaurea debeauxii also tend to be narrower and more ovoid (though this is not a very reliable characteristic) whereas this specimen has a nigra-like capitulum shape. Mixed populations are not uncommon where this was found (vc 15, Kent, UK). The flowering time (27 July) is compatible with this being either a hybrid or Centaurea debeauxii.
Above and below: Centaurea nigra
(flowering on the 22 July)
Above: Centaurea debeauxii flowering on the 7th august.
Further Complexities
Several different capitulum types occur in each species, but each individual plant has capitula of only one type. We have already seen radiate and eradiate forms, but other types are found. For example, in Centaurea jacea (Brown Knapweed) the basic type has hermaphrodite disc florets and sterile neuter ray florets, but there are plants with male capitula, in which the stigmas never separate so the florets are functionally male only. Some plants have female capitula with empty shrunken anthers and sterile marginal florets. Centaurea jacea is a species that usually has rays and the male capitula have the larger ray florets and present larger capitula. There is a whole spectrum between capiitula with all hermaphrodite (perfect) florets and those that are male only or female only.
Above: Centaurea debeauxii x nigra f. normalis.(It is hard to tell from a side-view whether this form is eradiate or semiradiate - it is useful to have bird's-eye views as well as side-views).
In the Centaurea nigra group or aggregate (C. nigra agg.), individuals of form longiflora (about 1 to 2% of individuals) have florets bearing some resemblance to ray florets with long petal tubes and these forms may or may not be radiate. On the other hand, form breviflora has short florets (short petal tubes) and in extreme cases the florets are all female, so this corresponds to the spectrum towards femaleness. Form normalis is the usual or basal type with hermaphrodite flowers.
Above: a form of Centaurea nigra with long disc florets (and also radiate / semiradiate); flowering on the 27 July. The slightly late flowering time suggests that some introgression from one of the other species is likely in this population. Radiant forms are also reportedly rare in 'pure' Centaurea nigra so I suspect there is some debeauxii genes in this population, indeed some individuals do score as hybrids in terms of phyllary overlap. Below: an eradiate form with short florets.
Above: an undeniably (pseudo)radiate form.
Above and below: these could be mistaken for short-flowered forms but they are actually going-over as their florets are withering. (They also seem to close before this happens and so bunch together, but this could also reflect the weather on that day). No pollen could be seen being presented, but a proper determination of gender would require dissection. Is this Centaurea nigra or C. debeauxii? See if you can decide by examining the third or fourth row of phyllaries from the bottom of the capitula.
In some populations the basic hermaphrodite types may occur alongside female individuals (derived by abortion of male organs, thus functionally female). This arrangement or plant breeding system is rare and is called gynodioecy and has important ecological consequences.
The population below is one of the mixed populations studied here.
Some of the forms flower early (in June) but most flower in mid to late
July onward, suggesting that this population is largely C. debeauxii,
but many of the individuals are hybrids and some are close enough to C.
nigra to be classified as such.
Flowering Times
Above and below: this is a mixed population containing a range of forms intermediate between Centaurea debeauxii and Centaurea nigra, but is probably mostly the former, taken on 27 July. (Click images to enlarge). Centaurea nigra mostly flowers in Britain in June, whilst C. debeauxii flowers about two months later in August. In this population a small number flowered in June, but most started flowering from the last week in July. Early observers stated that flowering time is a useful species diagnostic to distinguish these two. This is especially true in southeast England but further North and (and to some extent West) populations tend to flower later. M-J & T (1954) showed that flowering time is under genetic control, though affected by weather. First generation offspring between early and late flowering parents flowered at intermediate times (around July?). Some C. debeauxii have also been shown to flower early.
What about Centaurea jacea?
Centaurea jacea in its purest form is not as common in the British isles as it once was. This form was introduced from the continent and interbred with native Centaurea nigra and debeauxii to produce hybrids of varying degrees and even hybrids between all 3 species! However, it does still occur occasionally and evidence of its past presence can still be readily found.
Above: the brown phyllary appendages in this specimen may indicate introgression from Centaurea jacea (photo taken on 7 August at Polehill, Shorham, Kent, BI). The upper phyllaries are also very C. jacea like. The middle phyllaries, however, are typical of Centaurea debeauxii in form. Since we generally classify according to the middle phyllaries we would likely classify this specimen as Centaurea debeauxii even though it has indications of some Centaurea jacea gene introgression. Such populations with Centaurea jacea genes in them are apparently not uncommon on the chalk soils of southern England and these forms could be determined as Centaurea debeauxii x C. jacea or in some cases Centaurea nigra x C. jacea or the triple hybrid. All three hybrids are given the name Centaurea x gerstalaueri, Hybrid Knapweed.
Above and below: A form of Centaurea debeauxii with reddish phyllaries (with dark brown discs and light brown pectinations). The pale color of the florets indicates introgression from Centuarea jacea. (Photographed on 31 August at Queendown Warren, kent, BI).
Above: adults of Sloe Shieldbug (Dolycoris baccarum) congregating on a knapweed (7 August, Shoreham, Kent, BI). Note this is not a white-flowered form (though such do rarely occur) but rather a specimen that is going over.
Importantly, hybridization between Centaurea nigra and C. debeauxii is common where the populations overlap geographically. In Britain, Centaurea debeauxii prefers chalky soils in the South of England, whilst Centaurea nigra occurs more northerly and on clayey soils. Nevertheless, many populations in the South are mixed populations containing some hybrids as well as signs of introgression. (Introgression is when two different species hybridize and the hybrids backcross repeatedly with one parent form to produce a type that largely resembles either parent but may have a few features of the other species).
In practice the degree of introgression can not be absolutely determined on morphological grounds alone and there is a continuous spectrum of forms intermediate between C. nigra and C. debeauxii, but slightly arbitrary lines can be drawn between those forms intermediate enough to be considered hybrids and those close enough to either parent to be classified with it. Complications also arise as individual plants may lie in different parts of the spectrum within the same population and it is often easier to rate the degree of hybridity at the population level than for an individual specimen.
This calls into question the concept of 'species'. Animals, particularly vertebrates seldom hybridize (and if they do the hybrids are often totally sterile) and in this case it is usually sufficient to define a species as a group of individuals that can breed with one-another but not with other species. In asexual animals, a species is really a clone or group of individuals which are identical except for recent mutations in their DNA. In bacteria a different criterion is used since bacteria do not undergo sexual reproduction but can exchange packets of DNA with one-another. Here measures of the degree of genetic similarity are used: bacteria sharing most genes are common can be placed within a species.
A diploid Centaurea has difficulty crossing with a tetraploid Centaurea, because this would tend to produce a triploid zygote and odd numbers of each gene create problems during the production of gametes so triploid plants are often sterile (though in some genera of plants they are sometimes fertile, e.g. by doubling their chromosomes to form a hexaploid (6d) form). Indeed, a diploid Centaurea nigra can evidently breed more easily with a diploid Centaurea debeauxii than it can with a tetraploid Centaurea nigra.
Ploidy Levels
In plants things are rather more complicated! Plants also often exist as different cytotypes, for example in Centaurea nigra there are diploid and tetraploid forms. Diploidy (2n) is the norm in animals, in which each individual has two copies of every gene, one derived from the mother and one from the father. Tetraploidy (4n or 2d) occurs when an individual has twice the amount of genetic material and hence 4 copies of each gene, two paternal or two maternal. Male and female gametes are normally haploid in diploid species. Tetraploidy can arise in plants either when a fertilized egg (diploid) duplicates its DNA during development or when a diploid sperm fertilizes a diploid egg. Higher ploidy levels also occur in some plant groups. In the British Isles all the forms of this aggregate are apparently tetraploid.
In addition to different species, varieties and forms, there are different ploidy levels (chiefly diploids and tetraploids in this case) can exist within the same species but individuals of different ploidy levels do not interbreed so easily.
What about Centaurea pratensis?
Early accounts of the nigra group in the British Isles mentioned some additional species, one of which was Centaurea pratensis. C. E. Britton (1926), one of the early workers on British Centaurea stated that the species was 'not a difficult species to identify but there are in existence somewhat similar forms liable to be mistaken for it'. He explains that morphologically it is intermediate between C. jacea and C. nigra but that it should not be mistaken for the hybrid between these two species but that it may have been derived from hybridization. No evidence for this view is given, but the implication is that C. pratensis formed stable breeding populations as opposed to being a collection of recently produced hybrids. Under modern nomenclature it would most likely be classified as Centaurea jacea x C. nigra unless molecular evidence suggested otherwise. Centaurea pratensis is not currently recognized as a taxon in Britain.
What does DNA tell us about Centaurea?
The story of Centaurea pratensis does highlight the need for molecular studies to parallel morphological studies. A difference in morphology may or may not be enough to justify dividing a group into different species and likewise forms with very similar morphology can potentially have different evolutionary and genetic origins.
Molecular studies have been largely successful in sorting flowering plants into major groups such as families and in some cases genera. The relationship of the genus Centaurea to other Asteraceae genera and of the various sub-families of Asteraceae are fairly well established. However, genetic analysis has so far failed to adequately sort out the relationships between species in Centaurea. One problem is determining the species correctly prior to analysis. For example, Centaurea debauxii is sometimes classified as a subspecies of Centaurea nemoralis and so could be included as both Centaurea nemoralis, in the loose sense, and as Centaurea debeauxii in the same study.
However, the main problem is perhaps the fact that there is so much hybridization and introgression, resulting in horizontal gene transfer and what is sometimes termed reticulate evolution (like a tree with a web-like tangle of branches fused together at intervals, rather than as distinct branches). This can confound the problem of incomplete lineage sorting which results from multiple versions (alleles) of the same genes occurring within a population. This basically means that a phylogenetic tree (a tree showing the evolutionary history of related groups) based on a single gene may be inaccurate and depend on the choice of gene.These issues were addressed in Centaurea by López-Alvarado et al. (2014).
Typically, phylogenetic studies in plants make use of short regions of DNA, such as the ribosomal ITS (internal transcribed spacer) a region of DNA in the gene encoding ribosomal RNA (ribosomes are the protein factories of the cell) which is encoded on DNA in the main cell nucleus or short regions of the chloroplast DNA.Using such short segments of DNA López-Alvarado et al. obtained different trees depending which DNA region was used. Thus, molecular studies have not yet succeeded in sorting out the evolutionary relationships between Centaurea species. The potential solution is to incorporate larger regions of the DNA and better statistical methods are also being developed to generate more reliable evolutionary histories when there is appreciable incomplete lineage sorting.
What's in a name?
Centaurea includes plants commonly known as Knapweeds and Starthistles, the latter being more apt for those species with very spiny phyllaries. Centaurea nigra was previously called Centaurea obscura, due to the fact that the appendages of the phyllaries obscure the leafy parts of the phyllaries/bracts whilst 'nigra' means 'black' in reference to the appendages of the phyllaries usually being a dark brown or black color. It is worth noting that Centaurea nigra is divided into three subspecies by the Flora Europeae. Subspecies rivularis has radiate capitula, however, its appendages do not cover the bracts fully and it is found in parts of Portugal. Subspecies carpetana occurs in the West Pyrenees and mountains of central Spain and is eradiate with appendages covering the bracts. Finally subspecies nigra is also eradiate and the appendages also cover the bracts but this form occurs 'throughout the range of the species except Portugal'. However, in nigra the leaves are stated to be glabrous (hairless) whereas in carpetana they are shortly hairy with scabrid margins.
Stace (2014) refers to British radiate forms of C.nigra as subspeces rivularis and eradiate forms as subspecies nigra, but concludes that this is not a sufficient basis to designate them as separate subspecies. However, I have doubts that the subspecies seen on the Continent are the same subspecies seen in the British Isles: the described number of pectinations on the appendages seem to differ and radiate forms of C. nigra in the British Isles may well have acquired this trait from C. jacea introgression. Indeed, subspecies carpetana and rivularis could actually be hybrids.
Indeed, it is by no means certain that Centaurea nigra and C. debeauxii, as defined in continental Europe, are the same species as defined in the British Isles. Referring to type specimens may be of limited help if type specimens themselves show introgression. Several authors have noted that Centaurea debeauxii in the Britsih Isles (BI) is morphologically a more extreme form than the type specimens. In such a situation it might be better to chose new types from the extremes of the morphological spectrum. It should be noted, however, that introgression can potentially produce more extreme characters than are present in either parent, according to how the genes segregate and interact. For example hybridization of Sea Mayweed, Tripleurospermum maritimum with Scentless Mayweed, Tripleurospermum inodorum can result in (back-crossed) forms with more succulent leaves than were present in the maritime parent (Scentless Mayweed generally has less succulent leaves). Britton (1922) noted other differences between continental and British debeauxii and proposed designating the British form as form anglica.
Above: Centaurea debeauxii.
Centaurea surrejana ?
Britton (1922) recognized two additional species of the nigra group in the British Isles, neither of which is recognized today. The first of these is Centaurea surrejana. Its key features were branches that divide into short secondary branches which meant that the capitula were sometimes gemminate (occuring in pairs). It also had scattered (i.e. widely spaced out) leaves. According to M-J & T (1954) gemminate capituli indicates some level of hybridization with C. jacea (they produced one such form in a backcross between a C. jacea x C. nigra intermediate and C. jacea).
Centaurea drucei ?
The second species recognized by Britton was Centaurea drucei, which Lacaita (Mr C. C. Lacaita, another early botanist of Centaurea fame) found puzzling but noted was an intermediate between Centaurea nemoralis (i.e. Centaurea debeauxii) and Centaurea pratensis. Given that C. pratensis was intermediate between C. jacea and C.nigra, that potentially makes this form correspond with the triple hybrid: C. debeaxii x C. jacea x C.nigra.However, other specimens designated C. drucei appear to be Centaurea jacea x C. nigra and use of the name may not be morphologically consistent. The name may be a suitable one for suspected triple hybrids and the name may have been originally applied to populations with mixing between all three species.
Floret Color
Paler purple colors are usually associated with Centaurea jacea and its hybrids, according to M-J & T (1954). However, environmental factors can also affect floret color.
Above and below: Centaurea debeauxii x C. nigra. Of course cameras can lie about hues as this depends on lighting levels, which is why a study on color would have to include color standards, but some specimens in our hybrid population did seem by eye to have paler florets, even though they were growing in identical conditions. Note the uppermost phyllaries in C. nigra and C. debeauxii lack pectinations or fimbriae - the functional significance of this is explained below.
Above: stem leaf of C. debeauxii x C. nigra. The leaves are generally narrow with entire margins (and with a pair of basal stipules or stipule-like lobes in this one). Forms with stipuliform basal lobes have in the past been referred to as Centaurea drucei but are now generally considered a variant. A description of Centaurea drucei given by Britton sounds like a hybrid between C. debeauxii and C. nigra but the name has also been applied to various hybrids of C. jacea x C. nigra. The basal leaves tend to be larger, broader and more divided, as shown below:
Centaurea nigra and C. debeauxii are perennials. Some varities are recognised by some sources 9but some appear to refer to hybrids). For example, in Centaurea debeauxii there is a variety with less divided basal leaves in which the margin may be serrate but not as deeply lobed as the one above (var. subintegra). A dwarf variety which usually lacks branches and has small capitula has been called var. minima, but it is not clear whether this is a growth form or ecotype or whether the traits are inherited.
Below: Centaurea debeauxii growing on chalk meadow. Note the well-spaced and narrow phyllary appendages.
Above and below: Centaurea debeauxii. This population was still in flower at the end of August when most of the hybrid population (with intermediate flowering times) had gone to seed.
Below: Centaurea debeauxii is generally a less stocky plant than C. nigra with thin stems. It is often, though not always, quite elaborately branched in the upper parts.
Above and below: Centaurea debeauxii, this paler flower form was not uncommon but much less frequent than the more purple form. The petals of the florets are almost completely white (with traces of pink) but the anther tubes are still purple.
Seed Dispersal Mechanism
- seed shakers and elaiosomes
The fruit or achenes ('seeds') of Centaurea debeauxii and C. nigra (each consisting of a single seed enclosed by a thin and tightly fitting fruit wall with associated ornamentation) have a short pappus which may be very rudimentary and almost non-existent. Clearly such a pappus is not adapted for wind dispersal and it has been shown that Centaurea pappi of this type are effective at clinging to animal fur with the help of their short pappus filaments and hairs. However, in those with almost no pappus it is likely not specifically adapted for attachment in this way and may simply fall to the ground. It should be understood that sometimes short-range dispersal is a better strategy, replenishing populations growing in a suitable location. However, animals possibly disperse some achenes further afield to potentially new sites.
The greyish region of soft tissue to one side of the base, where the achene was attached during development, is derived from the parent plant and often misinterpreted as simply a scar of attachment or hilum. however, research has shown that it is a nutrient-rich elaiosome, for attracting ants which likely carry the achenes away and bury them to feed on the elaiosomes later (as has been definitely shown in some species). The achene is then left in a good place to germinate.
Above: a fruiting capitulum of Centaurea debeauxii in the open state and shedding achenes. The mechanism becomes clearer upon examination of the seeding flower-heads. The flowers detach, leaving the phyllaries in place, enclosing the achenes, each of which is surrounded by hairlike receptacle scales. The uppermost phyllaries lack pectinations or fimbriae and form a smooth rim or collar around the apical opening so as not to obstruct the achenes. The bottom is coated in a mat of hairs, which are the receptacle scales which surrounded and overtopped the developing achenes. Vibrations of the stem, such as may be caused by wind, or perhaps more likely a passing animal, cause achenes to be shed one at a time through the apical pore (which is usually not large enough to allow more than one achene through at a time though sometimes more cup-like. The mat of hairs keep detached achenes in the upper half of the chamber, queueing near the exit ready to be released when the structure is shaken again.
Furthermore, the phyllaries continue to undergo hygroscopic movements, opening when dry and closing in high humidity. This ensures the achenes can only be shed in dry weather. The mat of hairs may also help push the achenes up when the phyllaries close and press against them, but I have not verified this point.
Work in progress.