Umbel

This article is about the flower head of the carrot family. For the sunshade from which it takes its name, see Umbrella. For companies grouped under a single parent, see Holding company.

An umbel is an inflorescence in which the flower stalks, called rays, spring from a single point at the top of a stem, like the ribs of an umbrella. In a compound umbel each ray ends not in a flower but in a smaller umbel of its own, the umbellet. The arrangement is so characteristic of the carrot family that it gave the family its traditional name, Umbelliferae, “umbel-bearers”, which the rules of botanical nomenclature still allow as an alternative to Apiaceae.¹

In giant hogweed the terminal umbel of one Czech plant measured 85 cm across, and a single plant may carry nearly a hundred umbels of different sizes.²

The umbels of one plant are not equals. They form a hierarchy of up to four orders, in which umbels of higher order open first, carry a larger share of bisexual flowers and produce most of the seed, while those of the lowest order set almost none and serve as pollen donors.³

Etymology

Umbel entered English botany in the 1590s from Latin umbella, “parasol, sunshade”, a diminutive of umbra, “shade”. The same root gave umbrella, and umbrage, whose sense of offence came from the notion of being overshadowed by another and consigned to obscurity.⁴

Under a stand of giant hogweed the name is apt. Five centimetres above the ground, the share of daylight that gets through is on average 56 per cent lower than in uninvaded grassland.⁵overshadowed, and consigned to obscurity

Structure

                    .:*:*:*:*:*:*:*:*:*:.
                    |   terminal umbel  |   1 per plant
                    |     1st order     |   44.6% of the seed
                    '---------+---------'   flowers bisexual
                              |
             .----------------+----------------.
             |                                 |
      .:*:*:*:*:*:*:.                   .:*:*:*:*:*:*:.
      |  satellites |                   |   branches  |
      |  2nd order  |                   |  2nd order  |
      '------+------'                   '------+------'
       4 per plant                      3-4 per plant
       29.3%                            22.6%
             |                                 |
             '----------------+----------------'
                              |
                       .:*:*:*:*:*:*:.
                       |   tertiary  |   17 per plant
                       |  3rd order  |   3.5% of the seed
                       '------+------'   85% set no fruit
                              |
                       .:*:*:*:*:*:*:.
                       |  quaternary |   3 per plant
                       |  4th order  |   97.5% set no fruit;
                       '-------------'   pollen donors

Each compound umbel is made up of umbellets bearing many small, closely packed white flowers.² A large umbel has from 30 to 150 rays.⁶ The umbels of a flowering plant are ranked by the order of the shoot that bears them:²

• The terminal, or primary, umbel ends the main stem.
• Secondary umbels end the side shoots. Those that ring the terminal umbel are called satellites; those lower on the stem, branches.
• Tertiary umbels are borne on shoots from the secondary shoots, and quaternary umbels on shoots from the tertiary ones.

Vigorous plants may also send up shoots from the base of the stem. Their umbels rank between the first order and the second in size, in the number of seeds they set and in their proportion of male flowers.²

A typical flowering plant in the Czech study had one terminal umbel, four satellites, three or four branches, seventeen tertiary umbels and three quaternary ones.³ The total ranged from 5 to 98. Older floras give “usually 7–10 umbels”, evidently counting only the larger ones.⁷

Flowering

The terminal umbel opens first. The secondary umbels follow, and then the tertiary and quaternary umbels on the satellites and branches. Umbels of the same order open roughly, but not exactly, together.²

In western Bohemia flowering began between 20 and 27 June and peaked between late June and early July. A plant stays in flower for 36 days on average and for up to 60, the longer the more umbels it has. The terminal umbel is in flower for about ten days, and its fruit is ripe some six weeks after it opens.² Most seed is shed from late August to October.⁶

Protandry

Each flower is first male and then female, a condition called protandry. The five stamens shed their pollen within a day or two. Within an umbellet the flowers open from the edge towards the centre, and the first to open may then wait up to six days, neither giving pollen nor receiving it, until the stigmas of the whole umbel become receptive together, for a day or two. Each umbel thus passes in turn through a male phase and a female one.²

Protandry keeps a flower from pollinating itself, but not the plant. On 99 of 100 plants examined in the field, the male phase of some umbels overlapped with the female phase of others.²

Division of labour

Giant hogweed is andromonoecious: besides bisexual flowers, it bears male flowers, which shed pollen but have a reduced style or none and cannot set fruit. Where male flowers occur, they occupy the centre of the umbellets. Their share rises with the order of the umbel: the terminal umbel usually has only bisexual flowers, and the quaternary umbels usually only male ones.²

The result shows in the seed. Of 98 plants examined, none had a sterile terminal umbel, and in four-fifths of them more than three-quarters of its flowers set fruit. By contrast, 85 per cent of tertiary umbels and 97.5 per cent of quaternary umbels set no fruit at all.² The flowers of the lowest order act only as pollen donors.³to each according to its order

Bisexual flowers cost the plant more than male ones. Plants in the garden produced more all-male umbels than plants in the field, and the authors suggest that the stronger field plants, at a cooler and wetter site, could afford more bisexual flowers.²

Distribution of seed

An average plant in the Czech study set 20,671 fruits; the least fecund set 7,545 and the most 46,470.² They were shared among the orders as follows:

In a typical plant, then, the terminal umbel, one umbel in about 28, sets nearly half of the seed, while the tertiary umbels, three in every five, set 3.5 per cent between them. Seed from the terminal umbel is also the heaviest, about 16 mg against 11.7 mg from the satellites and branches. Heavier seed germinates faster, though not more often.⁸

Inequality

  share of the seed
 100% |                                       .#
      |                                     ..:#
      |                                   ..:::#
      |                                 ..::::##
      |                               ..::::::#
      |                             ..::::::::#
      |                           ..::::::::::#
      |                         ..::::::::::::#
      |                       ..::::::::::::::#
      |                     ..::::::::::::::::#
  50% |                   ..:::::::::::::::::##
      |                 ..:::::::::::::::::###
      |               ..::::::::::::::::::##
      |             ..:::::::::::::::::::##
      |           ..::::::::::::::::::::##
      |         ..:::::::::::::::::::::##
      |       ..::::::::::::::::::::::##
      |     ..:::::::::::::::::::::::##
      |   ..::::::::::::::::::::::::##
      | ..::::::::::::::::::::::#####
   0% |##########################
      +-----------------------------------------
       0%                50%                100%
       umbels, from the least productive

Plant ecologists have measured this kind of inequality since 1984, when Weiner and Solbrig proposed that the tools economists use for wealth and income, the Lorenz curve and the Gini coefficient, be applied to plant populations.⁹ The coefficient is 0 when every member has an equal share, and approaches 1 when a single member has everything. Applied to the umbels of a typical giant hogweed, it comes to about 0.8.^{[original research? – discuss]} The most recent estimates of income inequality give no country a coefficient above 0.6.¹⁰

Removal of the terminal umbel

Because the terminal umbel sets so much of the seed, it is the obvious target for cutting. Removing it alone achieves nothing: when the terminal umbel was cut off at peak flowering, the plant’s total output of seed did not fall.⁷ The loss at the top was made good by the rest of the plant. Even when every umbel was removed, nearly half of the plants regenerated before the season was out, producing on average about a hundred fruits each.¹¹ See hydra effect.a new head, the same body

Pollination

The flowers are open and unspecialised, with exposed nectar, and any insect that visits an umbel may pollinate it.³

At a site in Britain where giant hogweed grew beside the native common hogweed, 48 species of insect were caught on the giant and 59 on the native plant, and only 23 on both. Each insect carried mostly the pollen of the plant it was caught on. This selective foraging was judged one of the main barriers to crossing between the two species.¹²

A survey of twenty sites in the Czech Republic counted 2,611 visitors of 141 species on giant hogweed. Three-quarters of them were flies. The most numerous were the hoverfly Eristalis pertinax and the tachinid Gonia ornata, and among the bees and wasps, the honey bee. Native flowers at the same sites had fewer visitors in all, 2,181, but more kinds of them: 194 species. The hogweed’s visitors were numerous but few in kind and dominated by a handful of generalists, and the authors concluded that the plant is not a necessary resource for local flower visitors.¹³a mass market

Self-compatibility

Giant hogweed is fully self-compatible. In controlled crosses, umbellets pollinated by hand with the plant’s own pollen set as much fruit as those pollinated from another plant or left open to insects, some 88 to 92 fruits each. Umbellets bagged against insects set fewer than four.³ Insects are therefore needed for a normal crop, but a second plant is not: a single plant, isolated from all others, can found a population. This reproductive assurance is what Baker’s law expects of successful colonists.³

Crosses with the native common hogweed, Heracleum sphondylium, occur but are rare. The hybrids are intermediate in form and almost sterile.¹⁴ In experiments, crosses succeeded only with the native species as the seed parent. Unlike giant hogweed, most hybrids flower more than once.³

The hundred-thousand-seed figure

Figures of 100,000 seeds or more per plant are widely repeated. They appear to derive from counts of flowers.³

Sommier and Levier, who described the species in 1895, reported a plant at Geneva with at least 10,000 flowers. Later authors multiplied such counts by two, since each flower can give two seeds, and overlooked the male flowers, which give none. The largest plants counted by Tiley and Philp had up to 81,500 flowers, yet a plant with slightly more umbels set only 52,800 fruits. An often quoted figure of 120,000 seeds goes back to a note about 60,000 flowers.²³

Counted directly, the average plant set about 20,000 fruits and the most fecund 46,470. The authors consider 10,000 to 20,000 typical, about 50,000 an occasional maximum, and anything above 100,000 very doubtful.³ The larger figure continues to circulate. It doubled on its way through the literature without the plant producing a single additional seed.fictitious capital

See also

• Giant hogweed § Reproduction and spread
• Hydra effect
• Seed bank
• Andromonoecy, protandry and Gini coefficient in the glossary