Ecology largely consists in the search for natural patterns and the study of the underlying processes. Ecologists have often undertaken a further step by looking for generalities of these patterns and processes, in the search of the fundamental principles and general laws driving ecological systems. The search for generalities can be a great tool for improving our understanding of natural systems, particularly when it is accompanied by an explicit statement of hypotheses and expectations allowing them to be tested, often delimited, possibly falsified.
A possible drawback is represented by the temptation of using generalities to draw universal laws, which often turn into paradigms assumed to be true even if poorly tested ( or worse, proven false; Jeremy Fox calls them “zombie ideas” ).
The Bergmann’s Rule is a good example of the risks of over-generalization. In 1847, Bergmann suggested that, within genera of homeothermic animals, “smaller species would demand a warmer climate”; this was intended as a consequence of heat being more easily preserved by larger bodies because of their lower surface/volume ratio. As a consequence, Bergmann suggested that small species would be distributed at lower latitudes compared to larger congenerics. Bergmann’s original definition has been often revised ever since, more or less silently: some authors generalized it to cold-blooded vertebrates or even to all animals; others rephrased it to account for body size shifts along latitudinal gradients, without it being necessarily related to climate and, in general, without it being linked to any specific driving process; furthermore, the taxonomic resolution at which the Bergmann’s Rule was originally referred to (species within a genus) has been changed a number of times (see Blackburn et al. 1999 for a detailed review).
This continuous reinterpretation of the Bergmann’s rule made it equivocal: many studies claiming to confirm or confute it led to contrasting results, but their findings cannot be compared as they were testing different patterns under the same name; similarly, the pattern and the underlying process originally suggested by Bergmann have been mismatched, thus annulling the Rule’s original mechanistic explanatory ambitions.
Recently, two articles brougth some clarity and transparence to the issue: Daufresne et al.’s “Global warming benefits the small in aquatic ecosystems” (2009), and Adams et al.’s “Diatoms can be an important exception to temperature-size rules at species and community levels of organization” (2013). In both papers, Bergmann’s rule is still generally referred to all animals, thus maintaining some long-lasting confusion; on the other hand, it is explicitly distinguished from its intra-specific version, James’s rule, and from the Temperature-Size rule (TSR), that states specifically that individual body size of ectotherms decreases with increasing temperature.
Both papers proceed by clearly stating the possible underlying processes at the individual level, at the population level (decrease in mean body size) and/or at the community level (such as shifts toward small species) (see Fig. 1 in Daufresne et al.’s paper).
Daufresne and colleagues found the inverse correlation between temperature and body size widespread among different taxa but explained by different processes, while Adams et al. found that diatoms are one (more) taxon that likes breaking rules (of 31 species, 16 showed decreasing trends in body size and 15 showed the opposite pattern).
These two studies are a good example of how a fresh, more critical and more analytic view, together with a process-oriented approach, can help replacing reassuring but wrong paradigms with more sound and properly tested pieces of knowledge.
 According to Fox, they are concepts that “persist not just in spite of a single inconvenient fact, but in spite of repeated theoretical refutations and whole piles of contrary facts. They are not truly alive—because they are not true—but neither are they dead. They are undead”. http://dynamicecology.wordpress.com/?s=zombie+ideas&submit=Search