The conflicting evidence to date leaves open some
significant questions, all connected to the possibility
that the balance of stochastic and deterministic factors
influencing plant distributions depends on spatial scale.
First, how does the relative influence of the abiotic
environment or dispersal on the assembly of plant communities
change with the spatial scale sampled, especially
at local spatial scales? Second, to what degree
can the abiotic environment and dispersal limitation be
separated as independent factors affecting community
composition at different spatial scales?
In this study we investigated these questions about
the importance of dispersal and environmental determinants
of plant distributions. We extended our analysis
of previously published sampling data on ferns in
a forest understory, which were sampled at a mesoscale
(distances of 135–3515 m), in plots in which the correlation
between the geographic location and the abiotic
environment was eliminated (Gilbert and Lechowicz
2004). To contrast these mesoscale data, we sampled
ferns within a single hectare in the same forest,
with sampling sites in this fine-scale survey separated
by 4–134 m. These data allowed us to test the presence
and consistency of environmental and spatial patterns
within and between contrasting spatial scales. In addition,
the fine-scale survey allowed us to determine if
abiotic limits to plant distributions can complement
dispersal limitation, by examining the spatial distribution
of the environment at a spatial scale consistent
with dispersal from parent plants over only one generation.