Insect population ecology in British Columbia

In Volume 98, pp. 107-108, the historical account relating to the work of Neil Gilbert and collaborators was not correct. Neil wrote the following response: 'After cabbage aphid and Masonaphis, I thought it might be possible to introduce some limited generalization into population dynamics by constructing a Universal Aphid of which every aphid species would be a particular case. But pea aphid put paid to that. In the first two species, predation could be represented by a simple formula because it only took surplus production of aphids and was more or less compensated by density-dependent reproduction. In pea aphid, the coccinellids drove aphid numbers down low, so we either had to study predation in detail or admit defeat. It was all very messy but proved possible to predict predation rates, but not the numbers of beetles entering the field. That would be possible only if you knew the dynamics of the whole local ecosystem, an impossible task. In other words, the aphid population couid not be isolated from the rest, even to a first approximation. The whole thing was a failure as far as I was concerned, although a very instructive one.'