Nov 11, 2008
I can't remember who said that statistics will prove anything you want them to prove, but perhaps he was on to something. Whatever problem you happen to be working on, if you let a statistician near it, you will be given some glib solution that looks plausible but in most cases turns out to be quite worthless. Even worse is the reverence in which their gibberish is held by those who worship at the altar of science, without ever applying the simplest tests of common sense. When the Herald and TV3 can produce a political poll on the same day with the scores 10% apart, and each claims their survey has a margin of error of only 3%, it makes you wonder. How mad can you get?
The trouble is that statisticians do their calculations on the assumption that whatever they are looking for in a sample is evenly distributed throughout the material they are sampling. In fact, very few things are distributed like that.
Even starlings know better. If you watch a flock of starlings alight on a field in search of grass grubs or worms or whatever, they begin widely scattered, and as they search the soil in front of them, they keep one eye on their fellows. If some of the birds' heads change from a searching action to a feeding action, the outliers will move towards those ones, and this goes on until they are all concentrated in the two or three areas with the best yield of food. Follow your nose (or your beak).
We had a good example of the problem when the potato cyst nematode was discovered in Pukekohe in 1972. At that time, potato cyst nematode was the worst pest of potato crops in the world. It was the scourge of the industry in Europe, and had also gained a foothold in Canada and the United States, but was not known in Australia or here. There was no satisfactory control technique for it. You just had to avoid growing potatoes in fields that were infested with it. Its presence here could severely affect our potato crops and especially our exports to Australia.
The little cysts are full of eggs, and they can lie dormant in the soil for years. When a chemical released by a growing potato root reaches them, the tiny eggs hatch as microscopic worms which are attracted by the chemical and attach themselves to the fine roots. There they feed by withdrawing the sap. They also inject a fluid into the root which makes it swell and form a gall. By the time the worm is fully grown it has become so full of eggs that it assumes a spherical shape like a little white seed. When the root dies it becomes detached and forms a hard brown cyst that protects the eggs. Under a heavy attack the potato plant goes yellow and produces almost nothing. Anything that will move the soil will spread the cysts, so vehicles, machinery (rotary hoes, planters, diggers) or even footwear can carry them, as can the soil adhering to potatoes for sale or for planting. I remember telling a bunch of Maori ladies who were picking up spuds on an infested patch that they should dip their gumboots in a bucket of formalin before they went off the property. Some protested that they frequently went bare foot about the job. I relented and said it would be enough for them to wash their feet thoroughly in a bucket of water. One said "Sometimes we sit on the ground." She had me there.
Anyway it was necessary to survey the fields at Pukekohe and elsewhere to determine how far the problem was already spread. The technique for doing this in Europe and America, was to wait until the potatoes had been dug, and then to take soil samples, about a heaped tablespoon at a time, at measured intervals all over the field. These were then bulked together, mixed thoroughly, and a sample was processed so that any cysts in it would float to the surface where they could be collected and counted. Some statistician must have worked out the technique. The result told you something about the level of nematode infestation in the field. - but not much. For example it didn't tell you where in the field the cysts were concentrated, or indeed if your sampling had missed them altogether. After all, you had only examined abut a litre of soil out of the whole paddock.
Anyway, at Pukekohe potatoes are grown all year round, so at any one time, most of the fields are not empty. We surveyed the fields that were empty using the European method, but we needed some method for detecting infestations in fields where the crop was still growing. There were growers who wanted to know, before potatoes were dug, whether they could be exported, or sold on the open market, or carefully directed to some potato chip factory where the soil and peelings could be safely disposed of. And of course there were those in other parts of the country who wanted to know whether Pukekohe was alone with the problem, or whether they were to be quarantined too.
We decided to try the starlings' technique. When the little cysts are are still attached to the potato roots they are white and quite visible to the naked eye. A little research showed that if you dug a potato up at about 10 or 12 weeks of age, the white cysts would still be quite visible, and if you dug up a few potato plants you would know immediately if they were infested. By digging plants in ever increasing circles, you could also discover where the centre of infestation was, and the size of the infestation around it. Furthermore, since the chemical from the growing roots was actually bringing in nematodes from the surrounding soil, the plant was sampling a much greater volume of soil than the litre that the European method was. In fact it worked out to be about eighty times more effective.
When the little guy that drove the tractor at the Research Station saw some of our maps of infested fields, he could see the sense of them immediately. He had been a ploughing contractor earlier on, and when he saw the pattern he said, "Look! The contractor has come in here. He's late because contractors are always late, so he's done one round of the paddock to show the boss that he has arrived, and then he's sat down under a tree, and got out his bottle of tea. (Apparently vacuum flasks are too fragile and too expensive for contractors). Then he's cleaned the soil off his rotary hoe before he starts work on the rest of the paddock. If the last place he had ploughed was full of cysts, they get dumped there where he cleaned his gear. Each time the paddock gets ploughed in the following years, the cysts get spread a little, so you get this lens shape." Sometimes we could even identify the tree the contractor sat under with his tea bottle. These are the advantages of local knowledge and common sense over statistical sampling.
So we got to know quite quickly where the nematode was and where it wasn't, and the method was impressive enough that a visiting Australian nematologist was prepared to admit the import into Australia of potatoes from districts which we proclaimed to be clean. Wowee !
Sodium hypochlorite NaOCl, dip controls PCN on tubers, and breaks dormancy in early planting material.