Title:
======
  The Temporal Scales of Species-area Relationships


Abstract:
========
  The number of species living in an area increases with the
extent of the area. This relationship (termed SPAR) has been 
known to ecology for two centuries. Usually it fits a power 
equation. In the past decade, I have discovered that the pattern 
is really four patterns. The first depends on sampling bias: 
larger areas usually yield larger samples, and larger samples 
contain more species. The three other patterns exist at three 
different temporal scales characterized by rates that span six 
orders of magnitude: dispersal rates from generation to generation,
immigration rates to islands, and origination rates of new
species. The temporal scales correlate with spatial scales, but
are not identical to them.

The scale of a SPAR can be diagnosed by determining the power
that most closely fits it. Dispersal generates powers near 0.15;
immigration, powers between 0.25 and 0.55; speciation, powers
near unity. Ecology understands the rank order of these powers
but not their numerical values. These relationships should not be
reduced to a single equation because we know that causality is
reversed from the largest to the other scales, and because, in
order to apply species-area curves, it is important to preserve
the distinction among the biological processes that rule each
scale.

Although we do not know what abundance distribution characterizes
assemblages of species, we have a number of successful methods to
reduce most of the sampling bias. These now promise to allow us
to make fairly accurate estimates of diversity quite rapidly. We
can seek patterns in the fossil record with greater confidence,
and achieve more sensitive monitoring of diversity for
conservation purposes.

Understanding the differences among scales sheds light on many
ecological issues. It clarifies the co-evolution of habitat
differences. It predicts the echo pattern of local vs. regional
diversity. It predicts how the introduction of exotic species all
over the world will affect local and global steady states of
diversity. It also helps to evaluate applications of SPAR in
conservation.