The favourite books of Murray Lark
It is ironic, given his reputation as a scourge of religion, that I approached Richard Dawkins to tell him that I wished to abandon Zoology for Soil Science, with feelings one might ascribe to the eldest son of a staunchly protestant family informing the patriarch that he has signed up as a Jesuit. Dawkins was then my tutor at New College , Oxford , he was notoriously disdainful of applied science in general and agricultural sciences in particular. But I had just read the first book on my list: The Soil Under Shifting Cultivation by Peter Nye and Dennis Greenland. The book approached the practical problems of shifting cultivation in fundamental terms. How short can the cultivation cycle be if the system is to be sustainable? Nye and Greenland assembled data from trials in forests across the tropics, but this was not a prelude to a rather dull exercise in data mining. Rather they used the data to put reasonable bounds on parameters for a ?rst-order model of soil carbon. Having estimated these, they could explore a range of situations, showing how short the cycle could be before a long-term decline in fertility and stability would set in; rigorous concepts, mathematics and real problems. I knew then that soils science was for me! If I had to convince an eager school-leaver that his or her mathematical and scientific skills could be applied to problems both intellectually satisfying and important to their fellow humans, then I would give them this book.
The second book is Duchaufour s Pedology. I read Duchaufour in the summer vacation at the end of my second year. By then my head was buzzing with information about the soil. What Duchaufour offers is a system for this information. The book begins with an account of soil-forming processes and their function in the environment. Out of this emerges the concept of zonal and intrazonal soils, and so the ?nal chapters on the classes in Duchaufour s system. Again, the strength of this book is its conceptual approach. It would make excellent emergency treatment for a student whose passion for soil science is suffocating through enforced study of Soil Taxonomy.
These two books made me a soil scientist. The last one made me a pedometrician.
I arrived at university with an enthusiasm for statistics. None of the recommended texts gave quite the insight of Richard Webster s Quantitative and Numerical Methods in Soil Classi?cation and Survey. Despite the specialist title this sets out the principles of design-based statistics, with a bit more detail on multivariate methods, and a tantalizing half page on geostatistics, which Webster had only recently encountered when the book was published in 1977. I have recommended this book to students on introductory statistics courses, and those who need to get to grips with multivariate methods. In fact I nearly lost my copy when I lent it to a student who was puzzled by principal components. A second edition was published in 1990, with Margaret Oliver as coauthor: Statistical Methods in Soil and Land Resource Survey. It is a scandal that it is no longer available. Reprinting it would save us from much wasted effort through badly-designed sampling or unthinking use of statistical packages.