She loves me, she loves me not.

Pulling the ray florets off a daisy head one by one, while chanting the rhyme above, is supposed to help lovers to know if their love is reciprocated.  Clearly, odd numbers of rays will yield "she loves me", while even numbers will yield "she loves me not".  But ray number on daisies isn't random.  If you count the rays on heads of different species in the daisy family, you'll find some numbers keep cropping up again and again: 5, 8, 13, 21, 34, 55 (although individual heads might vary a bit around a mean that is one of these numbers).  Pick your daisy carefully and you'll get the answer you want.

She loves me, she loves me not, she loves me, ...

Leonardo of Pisa, a.k.a. Leonardo Fibonacci, was an Italian mathematician who introduced Arabic numerals to Europe in the early 13th century.  Every time you do arithmetic, remember it's because of Leonardo that you don't have to use Roman numerals (e.g., CXCVII + MDMLXVIII = MDCCXV).  He also introduced the number series that's named after him, originally demonstrated to predict the reproductive output of a pair of rabbits.  The Fibonacci series crops up again and again in biology, as in our daisy heads.  

In botany, one common place where we see it is in the arrangement of leaves on a stem, referred to as phyllotaxis. Leaves may be attached one at a node (alternate or spiral), two at a node (opposite), or three or more at a node (whorled).

Hebes, like this Veronica benthamii from Campbell Island, have opposite leaves, with each pair at right angles to the one below, an arrangement known as decussate.

Alternate leaves vary in how steeply they're spiralled.  When each leaf is directly above the one below (one complete turn between leaves), it is called monostichous, if it is at 180° to the one below (half a turn between leaves), so that they form 2 rows of alternate leaves, they are called distichous, and if they form an angle less than 180° they are referred to as spiral.

You can describe the spiral by finding two leaves that are one above the other. Now count the number of spirals around the stem between the two leaves, then the number of leaves in the spiral(s) (count the lower leaf of the pair as 0 and the number you’re after is the number of the upper leaf). You can then describe the phyllotaxis as a fraction, where the numerator is the number of turns of the spiral and the denominator is the number of leaves, eg, 2/5, 3/8.  

If you examine enough of these spirals, you’ll find the same numbers crop up again and again. These numbers are the Fibonacci series: 1,1,2,3,5,8,13,21,34...., which is made by adding two adjacent numbers to get the next one in the series. In plants, the phyllotaxis fractions that are found have the denominator as the next number but one in the series from the numerator, e.g., 2/5, 3/8, 5/13, 8/21, etc.

2/5 phyllotaxis

 Looking down on the stem from above, the angle between two adjacent leaves is given by the phyllotaxis fraction (1/1 gives 360° for monostichous leaves; 1/2 gives 180° for distichous leaves; 144° for 2/5 phyllotaxis. These fractions approach, but never exactly reach, 137° 30’ 28” of arc, and that’s the angle that results in the least amount of leaf shading along a stem. Finally, the really cool thing about these angles is they’re related to the Golden Ratio, widely used in Greek and Roman architecture and modern paper shapes[1]. If the angle between one leaf and the next is A, and B is 360–A, then the ratio of A:B is the same as the ratio of B:360.

Spiral phyllotaxis on a female cone of kauri (Agathis australis)

Many other spiral structures in nature follow the Fibonacci series, e.g., the arrangement of eyes on a potato or scales on a pine cone; it even describes the spiral of a gastropod shell. Note that a mathematical description, however elegant, is not an explanation. So why is the Fibonacci series so widespread in biology?  Spiral phyllotaxis according to the Fibonacci series allows these arrangements to keep the same shape regardless of their size, and this means the arrangement of the primordia at the tiny stem apex is just the same as the arrangement of the mature leaves on a thickened stem.

---

[1] The A series of paper sizes (e.g., A4) is organized on the golden rectangle.  Thus A5 is half the area of A4, but exactly the same shape, and is obtained by cutting A4 in half cross-wise.  A0 has an area of 1m².

Not tonight Josephine.

Apparently Napoleon never said "Not tonight Josephine", but it's probably the one thing most people associate with his wife, the Empress Josephine of France (1763–1814).

She should be more widely known instead for having this spectacular climbing plant named in her honour:

Copihue, Lapageria rosea

The Empress was born Josephine Tascher de la Pagerie, and it's from her surname that Lapageria is derived.

Lapageria rosea is the only species, and it's the national flower of Chile, which is where it occurs naturally.  I've been watching this plant in Kelburn, since noting its spectacular flowers.  The flowers are clearly monocot flowers—they have sepals and petals in threes—but take a look at the leaves:

Lapageria leaf, and note the Agapanthus leaf behind, a more typical monocot.

Monocots are supposed to have parallel leaf veins, and these are clearly reticulate (net-like), which is generally characteristic of eudicots and basal angiosperms.  What's going on here?

Judd et al. (2008) explain: "Several monocots have pinnate to palmate leaves with obviously reticulate venation patterns ..., but these are probably reversals associated with life in shaded forest understory habitats".  How do they know this?

There are two possible ways that some monocots could have net-like leaf veins.  First, the groups with net-like veins could be at the base of the monocot family tree, and then we'd interpret their venation as an ancestral dicot-like feature that they've retained. 

Alternatively, they could be derived higher in the family tree of monocots, from ancestors that had parallel veins, in which case we'd infer that there was a shift from net-like to parallel at the base of the monocots, followed by a reversal to net-like higher in the tree.

When we look at a family tree (a phylogeny) of the monocots, it's clear that the second interpretation is the correct one, because these net-veined plants are deeply nested among species with parallel venation.  Indeed, it seems this reversal has happened more than once.  Lapageria is classified in family Philesiaceae among the Order Liliales.  Some other net-veined monocots are also placed in Liliales, such as our native Luzuriaga (Alstroemeriaceae) and supplejack, Ripogonum  (Ripogonaceae).  But net-veined Dioscorea (the true yam, not to be confused with oca, Oxalis tuberosa) is classified in Dioscoreales, so it looks as if they came by their net-veined leaves as a separate event.

Luzuriaga parviflora

It seems the genetic ability to make net-veined leaves might be present in many monocots but silenced in some way, making it relatively easy for reversals to evolve.  There are several well-known mechanisms for this kind of evolution.  (Added 8 July 2011: Here's a great description of how this kind of thing can work)

Lapageria rosea (local name copihue) is the national flower of Chile.  Hanging red flowers from South America, with no landing place for a pollinating bird, are characteristic of hummingbird pollination.  Sadly, we have no hummingbirds in New Zealand.  The plant I've been watching does set fruit, but the fruits seem to fall before the seeds are ripe and it may be that fruits begin to develop even when the flowers haven't been pollinated.  The anthers are close to the stigma, so you'd think it could self-pollinate, but maybe some other mechanism is preventing seed set.  I might put on my bumble-bee suit and try pollinating it by hand.

Inside the flower, showing the 6 anthers close to the stigma.

Reference

Judd, WS; Campbell, CS; Kellogg, EA; Stevens, PF; Donoghue, MJ (2008).  Plant systematics a phylogenetic approach (3rd ed.).  Sinauer, Sunderland MA.

A walk in the Waitakere Ranges

For a Wellingtonian, it's amazing to get into the forest in the north of New Zealand where there's much more diversity.  This post is just a set of photos from a nice afternoon walk.  Many familiar plants are here of course, but some have larger leaves or other small quantitative differences.

A small kauri, Agathis australis, about 2m dbh.

On the ridges, the dominant tree is kauri, Agathis australis.  There are other conifers too, totara (Podocarpus totara), rimu (Dacrydium cupressinum), and kahikatea (Dacrycarpus dacrydioides).

Leucopogon fasciculatus, (Ericaceae)

Metrosideros perforata (Myrtaceae)

New Zealand's only palm, Rhopalostylis sapida, is common in the understory.

Tree ferns are common in the understory too.

Umbrella moss

Veronica macrocarpa, a northern hebe

Seedlings of kahikatea, Dacrycarpus dacrydioides.

A small pukatea, Laurelia novae-zelandiae.

Pukatea has spreading buttress roots that help support the tree in its swampy habitats.

Lobelia physaloides

It's well known that New Zealand native flowers are, frankly, a bit boring.  Most are small, and white, green, or yellow.  There are a few red ones, like rata, pohutukawa, and kakabeak, associated with pollination by birds.  Blue flowers are very rare.

When it comes to our fleshy fruits, many are small and red or orange, associated with dispersal by birds.  There are small blue, purple or white fruits too, and some translucent ones.  Large fruits are quite rare, probably because our fruit-eating birds can't swallow anything larger than a karaka or tawa fruit.

That all serves to introduce the exceptional Lobelia physaloides, one of the most striking plants in the New Zealand flora.  Its flowers and its berries are large and purple.  The berries have fine white stubbly hairs on ridges running lengthwise, and at the apex where the remnants of the calyx are found.

Lobelia physaloides fruit; it's about 18 mm long.

This plant has usually been called Colensoa physaloides, or sometimes Pratia physaloides.  Botanists generally are adopting a wider view of Lobelia, which includes Pratia, Isotoma, and Hypsela.  But L. physaloides is so unlike the little creeping New Zealand Lobelia that many people still want to treat it as a separate genus.  How reasonable is this position?

First, there's good evidence L. physaloides isn't closely related to the other New Zealand species of Lobelia.  But is it still a species of Lobelia anyway?  This is one of those situations where looking at our flora in isolation from the rest of the world isn't a good idea. Overseas, many lobelias are large bushy plants with coloured flowers, some, like the spectacular species on African mountains or in the Hawaiian Islands, are small trees.

Antonelli (2008) showed L. physaloides belongs close to some African species; together they're sister to the rest of the genus in the expanded sense.   Lammers (2011) followed that up with a formal taxonomic treatment, and he puts L. physaloides on its own in its own section (sect. Colensoa) sister to a large section of species that are found in Africa and Asia.  If Colensoa is to be recognised as a genus, its large sister group would either have to be included in Colensoa too, or would have to be recognised as another genus.  These botanists with a global view haven't split Lobelia, and it seems New Zealand botanists might be influenced more by the differences than the relationships in this case, although they've been happy to include Pratia, Isotoma, and Hypsela in a broad circumscription of Lobelia.  

Divergent growth forms, flowers, and fruits in Lobelia seem to have evolved convergently many times.  Using such differences as the basis of a classification might risk putting unrelated plants together in a genus, or at least classifying plants in separate genera from their nearest relatives.  

Sectioned fruit of Lobelia physaloides.

More interesting is the question of what pollinates the flowers and disperses the seeds of L. physaloides.  Inside, the fruits are hollow and a bit like little bell peppers, with thousands of tiny seeds on a pithy white placenta.

What eats the fruits and what disperses the seeds?  Candidates might include birds, weta, and lizards.

References.

Antonelli, A. 2008: Higher level phylogeny and evolutionary trends in Campanulaceae subfam. Lobelioideae: Molecular signal overshadows morphology. Molecular Phylogenetics and Evolution 46: 1–18

Lammers, T.G. 2011: Revision of the infrageneric classification of Lobelia L. (Campanulaceae: Lobelioideae).  Annals of the Missouri Botanical Garden 98(1): 37–62.

Seasons and antipodes

I like to walk from Wellington city to the VUW Kelburn campus via Salamanca Road for two reasons.  First, if you could drill straight down through the Earth, you'd come out in a field in Spain just off the A62 highway, where the nearest town is Salamanca.  That's cool.  Secondly, on (rare) still sunny winter days like today you get to walk past a flowering wintersweet in someone's front garden.

You smell it before you see it, a heavy sweet perfume; look around and there it is, just over the front fence.  Chimonanthus praecox, wintersweet, is a native of China and a member of the basal angiosperms.  That is, although it's a flowering plant, it's neither a monocot nor a eudicot.  It's classified in the family Calycanthaceae and the order Laurales, in the Magnoliid complex.  There's a clue to that: the twigs smell peppery from ethereal oils, a group of chemical compounds that are characteristic of many Magnoliids and some other basal angiosperms.  These oils give us spices like pepper, nutmeg, and cinnamon.  If you're in Wellington, take a walk past and have a sniff.  Enjoy, but please don't pick any; I'd hate to be the cause of the shrub being ravaged by my horde of readers.

Jovellana

Jovellana sinclairii

Jovellana sinclairii is one of our prettiest New Zealand native plants, but it's not well known.  The picture above was taken with a flatbed scanner, so it looks a little unnatural, but it shows the detail of the flowers.  Here's J. sinclairii au naturel on a roadside bank at Tiniroto on the east coast of the North Island:

Jovellana sinclairii, Tiniroto

We have two species here; the other one is J. repens, which is much smaller:

Jovellana repens, from Akatarawa, Wellington

Jovellanas seem quite easy to grow.  I have both species in pots, and they flower well.  I'm hoping to plant them out in a shady spot soon.   Other species of Jovellana are native to South America.  They're closely related to the amazing slipper flowers, Calceolaria, also South American.   Jovellana and Calceolaria are nowadays classified in the small family Calceolariaceae.

Calceolaria tripartita, Wellington

C. tripartita seems to be becoming more common as a minor weed around Wellington; this one was on a roadside bank opposite the Botanic Gardens.

In 2009 Geoff Davidson, Peter de Lange, and I described a new species of Veronica from just north of Auckland, calling it Veronica jovellanoides because of its resemblance to Jovellana repens.

Veronica jovellanoides.

Reference.

Davidson, G. R.; de Lange, P. J.; Garnock-Jones, P. J.  2009: Two additional indigenous species of Veronica (Plantaginaceae) from northern New Zealand: V. jovellanoides, a new and highly endangered species, and V. plebeia R.Br.  New Zealand Journal of Botany 47: 271–279.

Casting Nasturtiums

Nasturtiums are a source of a little confusion in the botanical and horticultural worlds.

The trouble is, there's more than one plant that goes under this name.  The name nasturtium is derived from the Latin for "twisted nose", a reference to the strong smells and tastes of these plants.  These characteristic smells and tastes come from glucosinolates, or mustard oils.  Glucosinolates are a family of compounds that contain sulphur and nitrogen; their main role is the deterrence of grazers.  However, some insects (cabbage aphid, cabbage white caterpillar) are able to either break down or sequester the glucosinolates and have the cabbage resource all to themselves.

Watercress growing in a stream in Wellington

First, a plant that has Nasturtium as its scientific name.  This is the watercress, a member of the mustard family, Brassicaceae.  For a long time there was thought to be just one species, Nasturtium officinale.  The advent of chromosome counting turned up a surprise: some plants had four sets of chromosomes instead of the usual two, and they also had small but consistent differences in flower size, seed ornamentation, and arrangement of seeds in the pods.  A new species of watercress, N. microphyllum, was named.  Hybrids between them are mostly sterile.

Later, it was decided that Nasturtium as a genus wasn't very different from Rorippa, and the two genera were merged under the latter name.  Our two watercresses are now called Rorippa nasturtium-aquaticum and R. microphylla.  The glucosinolates of watercress are quite mild and make them a nice addition to salads and sandwiches; some of the native New Zealand bitter-cresses (Cardamine) have similar flavours.

Garden nasturtium, Tropaeolum majus.

What about the garden nasturtium?  This one has nasturtium as its common name, hence the confusion. Its scientific name is Tropaeolum majus, and it's in a different family (Tropaeolaceae) from watercress.  Its leaves are a great addition to salads, and their flavour is a bit like capers (Capparis, family Capparaceae). In fact, glucosinolate producing plants are found in about 8 or 10 families.  These families look so different from each other that it was assumed glucosinolates must have evolved independently many times in unrelated groups.  For instance glucosinolates are found in the Cleome family (Cleomaceae), the mignonette family (Resedaceae), and the papaya family (Caricaceae).  It seems morphology was telling us these families weren't related, while chemistry was telling us they were.

Spider flower, Cleome hassleriana.

How can botanists sort out a conflict like this?  Doesn't it threaten our notions of evolution?  Well, no.

One approach, but not a rigorous one, is simply to privilege (to borrow a verbed noun used in the humanities) one set of data over another.  That's how we mostly dealt with the glucosinolate problem; we assumed that morphology trumped chemistry, and that glucosinolates had arisen many times independently in mostly unrelated families.  Lazy, and wrong, as it turns out.

Mignonette, Reseda lutea.

First, when data seem to conflict like this, we need to look at the differences and similarities and see if there are developmental differences among some of the similarities, or underlying structural similarity in some of the differences.

Another approach is to find a third and independent set of data.  That came along in the 1990s with the huge DNA sequencing effort in plant taxonomy.  And here was a genuine surprise (Rodman et al. 1993): nearly all the glucosinolate families are closely related, because the DNA data agree with the chemical data.   We need to re-assess the morphological differences.

It seems glucosinolates have arisen just twice: once in a large group of families (now grouped together as Order Brassicales, e.g., Brassicaceae, Capparaceae, Tropeolaceae, Caricaceae, Resedaceae) and another time in the unrelated small genus Drypetes (Euphorbiaceae).  Future study will focus on the huge range of different growth forms, flowers, fruits, and seeds in the Brassicales.  It's going to be interesting as the genes that control this development are discovered and their function becomes understood.

Reference
Rodman, JE; Price, RA; Karol, K; Conti, E; Sytsma, KJ; Palmer, JD 1993:  Nucleotide sequences of the rbcL gene indicate monophyly of mustard oil plants.  Annals of the Missouri Botanical Garden 80: 686–699.

Sweet alyssum, Lobularia maritima (Brassicaceae)

How can a flower like this exist?

A couple of years ago we had a visiting class of students from Lewis & Clark College in Oregon for a semester of New Zealand studies.  They were bright, friendly, and fun to be with as we travelled the length of New Zealand and even to the Subantarctic islands.  I’ve enjoyed keeping in touch with them through Facebook as they graduate and spread out over the USA beginning their careers and further study.

Lewis & Clark group on Campbell Island

Last week one of these students posted a photo on Facebook of this amazing little flower, the steer’s head (Dicentra uniflora).  It’s a spring-flowering plant of British Columbia and western USA, often seen in Yosemite and such places.  She asked, “How can a flower like this exist?” and I thought, “What a great question for a blog entry”.

There are many answers to the question, depending on what is really being asked.  First, these students aren’t credulous creationists, so this wasn’t the standard statement disguised as a question, suggesting that something so wonderful couldn’t possibly have evolved.

A first answer might look at the question of how the flower functions.  We can be pretty sure it’s not actually mimicking a steers head, but in that part of the world, it’s a natural thing to see the likeness, a kind of pareidolia again.  The key to the function question is, “What pollinates it?”  There’s a thesis by Julia Rayl at Washington State University on just this topic, but it appears not to be available on line.  However, a published guide to selecting plants for pollinators lists bees as the pollinators of the steer's head.  That’s consistent with other American species of Dicentra, which are pollinated by queen bumble bees (Bombus).  It’s also what I would have guessed from its morphology: a closed flower that needs to be forced open by a visitor, with a nectar-spur.  So the form of this flower works to attract bees and probably to make it difficult for other insects to steal nectar or pollen.

Corn poppy (Papaver rhoeas), ramping fumitory (Fumaria capreolata), wall fumitory (F. muralis)

Another way to answer the question is to ask how it evolved.  Dicentra is a member of the family Fumariaceae, which all have flowers a bit like this.  Fumariaceae are close relatives of the poppies, Papaveraceae; in fact many botanists treat these two as a single large family.  The key differences are in the flowers and the sap.  Poppies have a milky latex that becomes sticky and finally solid as it dries; the dried latex of Papaver somniferum is opium.  In Fumariaceae the juice is watery.  Poppy flowers (above, left) are open and symmetrical with many separate stamens, whereas most Fumariaceae flowers (above, centre) are closed to form a tube-like structure, with a pair of branched stamens (that look like bundles of three) on each side of the flower.  Both families have flowers with four petals—not a common number in the flowering plants—considered to be presented in two opposite pairs.  In poppies the outer pair and the inner pair are much the same, whereas in Fumariaceae they're different and one or both outer petals has a spur.  In Dicentra both outer petals are spurred, and in Fumaria, only one is.  The evolutionary history of the poppies and fumitories can be shown in a branching family tree diagram, or cladogram:

This shows the common ancestor of Dicentra and Fumaria, and a bunch of other genera, was the sister species to the common ancestor of the poppies (represented here by Eschscholzia and Chelidonium).  The diagram was developed by comparing the changes in DNA sequences among these plants, and it's a bit simplified from the original.  We can use the order of branching to make some inferences about the likely order of change in flower features, by mapping the flower features onto this diagram.

From top, Californian poppy (Eschscholzia californica), steer's head (Dicentra uniflora), ramping fumitory (Fumaria capreolata)

It looks like the first evolutionary step was the development of an asymmetrical flower with two short nectar spurs, like that found in Hypecoum, sister to the rest of the Fumariaceae.  Next, the flower has become closed in the ancestor of Dicentra, Fumaria, and their relatives.  The curling back of the outer petals is accentuated in the steer’s head flower, Dicentra uniflora, but it’s seen in other species of Dicentra too, and related flowers, like Lamprocapnos spectabilis. A later step, after the Dicentra lineage diverged, involved loss of one nectar spur to make a flower with only one plane of symmetry, typical of Corydalis and Fumaria.  

The key to understanding this is the realization that evolution of complex structures occurs step-wise, with each step being adaptive.  Richard Dawkins explains it well in Climbing Mount Improbable. 

A third way to answer is to ask about what’s going on as the flower develops.  In other words, how do the evolved differences at the gene level affect flower development to actually change the flowers as they grow?  Nowadays, botanists are homing in on the actual genes that are responsible for features like floral asymmetry and changes in the timing of development, and early results in Dicentra and Fumaria are promising, even though they haven’t yet fully answered these questions (Kölsch & Gleissberg 2006).

In short, we can explain how the flowers work to attract their pollinators, the series of evolutionary steps that led to this flower as well as the order they occurred in, and I’m confident that very soon the mechanisms of development of the steer’s head flower will be known.

References

Kölsch A; Gleissberg S (2006).  Diversification of CYCLOIDEA-like TCP Genes in the Basal Eudicot Families Fumariaceae and Papaveraceae s.str.  Plant Biology 8: 680 – 687

Pareidolia in the lunchbox

The term Pareidolia refers to how we think we see a shape in a random pattern: a cloud looks like an animal, or a slice of toast has an image of Jesus.



This kiwifruit was in my lunchbox today.

Kiwifruit (Actinidia deliciosa; never call it kiwi) is a fruit formed from a compound ovary, with many individual carpels.  In this fruit, it looks like one carpel didn't develop.  By remaining small and empty of pulp and seeds, it held the fruit's wall close to the central core, while the other carpels grew and swelled around it.

In this cross section you can see the walls of the individual carpels, 31 of them by my count.

Sometimes, when things go wrong with development, it helps us understand the processes that are normal.