or inclusive niches [31,33,35–38] (Figure 1d), where most
species perform best in benign, productive sites (i.e.
undisturbed sites with a higher concentration of
resources). However, the ability to dominate productive
sites or to sequester high-quality resources trades off with
the ability to persist on low-quality resources or to tolerate
harsh conditions. For example, a species can allocate
resources to either frost tolerance or growth rate [39] and,
similarly, a desert rodent can have either predator escape
mechanisms (tolerance) or competitive dominance [40]
(Box 1). In some cases, there might be a mixture of the two;
animals might have shared preferences for habitat and
distinct preferences for food or the appropriate model
might change with scale. The model of shared preference
has been developed repeatedly and shared preferences
appear to be twice as common as distinct preferences [33],
yet there is a continuing emphasis on distinct preferences,
probably because most community theory embraces
species interactions and remains vague about environmental
gradients. Traits and performance currencies
could make the above verbal theory more quantitative
and rigorous (Box 2).
Prioritizing factors
Many factors (e.g. traits, environmental variables and
performance currencies) could influence community
structure. However, it is impossible to treat every factor
equally and to study each variable and all interaction
terms simultaneously. Even three traits and three
environmental variables, each studied at five levels,
would require a prohibitive number of measurements
(53C3Z15 625 treatments in a complete design). Factors
must thus be prioritized using existing knowledge of the
study system (Box 1). Without claiming that lower ranked
traits and abiotic factors have no effect, we expect them to
have less predictive power than do the higher ranked. For
fundamental niches, the highest-ranked traits will interact
strongly with physical gradients in influencing the