Читать книгу Forest Ecology - Dan Binkley - Страница 14

The most important point is one that is not found in the graph, but applies to this graph and most others in this book. Graphs plot the values for a variable (such as forest growth) based on another variable (such as precipitation). Even when the association between the two variables is very strong, it's fundamentally important to recognize that evidence of an association is not evidence of a cause‐and‐effect relationship. The forests that provided the data for Figure B had very different species composition, different soils, different ages, and different local histories of events. Some of these may happen to vary with precipitation, and might be the actual drivers of the trends that relate to precipitation. Similarly, if forest growth tended to decline in the warmest sites, that might result from increased activities of insects (or monkeys) rather than a direct effect of temperature.

Identification of driving causes behind patterns requires other sorts of evidence, especially evidence from experiments. If the addition (or removal) of water changed growth as much as was expected from the geographic gradient, then increased confidence would be warranted in water influencing growth across many locations. If plantations of a single species also declined in growth at high temperatures, then the trend in Figure B may be less influenced by changes in tree species across sites.

This fundamental idea is summarized in the aphorism, “Correlation does not equal causation.” All scientists know this, but placing science into sentences can be challenging for both thinking processes and writing processes. It's easy to find examples where scientists forgot this basic point (perhaps even a few places in this book?).