Rothamsted

Rothamsted Manor

I had indirect information about Rothamsted Experimental Station long before I visited or learned about the real thing. In Aldous Huxley’s utopian novel Island, which I read while a student in Australia, the island of Pala has an experimental station called “Rothamsted-in-the-Tropics”. Also, the railway halt nearest to Gatton Agricultural College was named Lawes, after Sir John Bennett Lawes, First Baronet Lawes (1814-1900), scientist and founder of Rothamstead Agricultural Experimental Station in Hertfordshire, developer of the use of artificial fertilizers and particularly superphosphate. (From the Queensland Government placenames website.)

(I passed Lawes on the train journey between Brisbane, where I studied, and Toowoomba, where I lived, on a number of occasions. By then, the train only ran to Helidon at the foot of the range, where it was met by a McCafferty’s bus, so I missed the spectacular journey up the range which I had done as a child. Later the bus went all the way to Ipswich to meet the train; now the entire journey is by bus, and the McCafferty’s buses have been re-branded as Greyhound.)

Anyway, as the Wikipedia article quoted above says, the experimental station was founded by John Lawes, a pioneering manufacturer of artificial fertiliser. Of course it was not for disinterested research: Lawes wished to know if the use of artifical fertiliser could obviate the need for crop rotation and fallow periods, so in 1843 he set up an experiment on Broadbalk which is still running. The field was divided into strips which were given different fertiliser regimes and cropped continuously with winter wheat.

Another classic experiment which is still running is Park Grass, where the effect of different fertiliser regimes on the hay production of grassland was studied. This experiment was founded in 1856. Nowadays, it is regarded more as an ecological experiment (with undisturbed plant communities whch have been subject to continuous study) than an agricultural one.

In both cases, there have been some changes to the experimental management over a century and a half. For example, on Park Grass, it was found that some forms of nitrogen fertiliser caused soil acidification, so the strips were divided and different levels of lime were applied to combat this.

The picture shows the boundary between adjacent plots in Park Grass. The plant community on the left is much less diverse than that on the right, an effect of nitrogen fertiliser, which encourages plants that respond well to it to take over the plot.

Park Grass

The third afternoon of the Designed Experiments: Recent Advancess in Methods and Applications workshop at the Isaac Newton Institute was devoted to a trip to Rothamsted. The Research Station is, more-or-less, holy ground to statisticians, especially those working in experimental design, because R. A. Fisher worked there for many years. He introduced balanced incomplete-block designs and Latin squares, previously recreational curiosities, into agricltural research; but, much more significantly, he developed three principles that underlie the subject: replication, randomization, and local control (blocking), which with better reason than the traditional use of the term are called the “three Rs” of experimental design.

There were some problems. Not enough time had been allowed for the coach trip from Cambridge, and so we arrived rather late and the visit had to be curtailed; but we did get to see the Broadbalk and Park Grass experiments. Some of the statisticians were horrified by the lack of replication and randomization in these experiments; of course, they were laid out long before Fisher’s work. There is some semblance of blocking now that the plots have been divided, with some treatments applied to whole plots and others to subplots (actually this is a split plot design rather than a block design).

At the end of the visit, we were treated to a wonderful lecture by Roger Payne on the history of statistics from a Rothamsted point of view. Roger claimed, with documentation, that many important advances in statistics had come from Rothamsted, courtesy of Fisher or his successors such as Frank Yates, John Nelder, John Gower, my colleagues Rosemary Bailey and Donald Preece, and Roger himself, all of whom worked there.

Something else that horrified the non-Rothamsted people was that papers had to be approved by somebody upstairs before they could be submitted for publication. But I think this is not so uncommon outside universities, and it may well be the future for us as well: at least, the choice of journal may be dictated to us.

The slides of Roger’s talk, including some extremely interesting quotations, should appear on the DEMA website soon. I will put in a pointer when this happens.

Advertisements

About Peter Cameron

I count all the things that need to be counted.
This entry was posted in geography, history and tagged , , . Bookmark the permalink.

One Response to Rothamsted

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s