garlick’s notebook / By Don Garlick

A Counting Problem

(Dec. 27, 2007)  This Holiday season deserves a fun project: Count the seeds in one cattail. I estimated the number to be half a million, but every scientific claim should be verified before acceptance. I suggest you share the fun and check my result. A source of cattails is just east of Bigfoot Gas in McKinleyville. The cluster of seeds that I examined was a cylinder 20cm long with a circumference of 9cm. Its area was thus about 18,000 square mm. I attempted four methods of obtaining a sample small enough to conveniently count:

In one square millimeter of the surface I counted 20 remnant stigmas (pollen traps). This yielded an estimate of 20x18,000 = 360,000 seeds.

GALLERY >

I removed a short length of the cylinder by means of two closely spaced cuts. A photo shows why this method failed. The first cut allowed seeds to expand out of control so that the second cut included many trespassers.

I marked a square centimeter and pulled the seeds from that area. I divided these into 4 equal piles and repeated to yield a 1/16 fraction which was countable. This method yielded an estimate of 640,000 seeds.

I used a drop of molten candle wax to bind a clot of seeds which I then extracted. This yielded an estimate of 450,000, calculated as follows: Diameter of clot = 4mm. Area = pi x 2mm squared = 13 square mm. I counted 65 seeds in one-fifth of the clot. Thus total = 65x5x18,000/13 = 450,000 seeds.

The average of my estimates is 460,000 with a large uncertainty, but I will tentatively claim that one seed cluster can contain half a million seeds. These plants usually produce more than one cluster and often regenerate vegetatively from rhizomes, so the significance of my counting is that, in a stable population, one seed has less than one chance in half a million of growing to maturity.

I would like to hear from you, but realize that I am more likely to hear from whoever has to clean up a mess of seeds drifting in the slightest breeze.

1 SHARE