Ten paired plots were established on horizontal rock faces near the Ben Lily Memorial in the Gila National Forest NM in 1978. All 10 cm x 10 cm plots were similar in slope, elevation (2100 m), and light exposure. One of each pair of plots was randomly

1
Testing mechanisms of the intermediate
disturbance hypothesis using long-term data of
saxicolous lichens
A. Pastore, C. Prather, E. Gornish, R. Ellis, and
T. Miller Department of Biological Science,
Florida State University, Tallahassee, FL 32306
For more information on this and other thrilling
projects, please contact Abigail Pastore at
apastore_at_bio.fsu.edu
Results
Introduction
1980
1985
1988
Figure 4. Changes in lichen cover through time on
a single plot.
Figure 2. A) Rates of colonization for the six
most abundant species on the disturbed plots. B)
Competitive abilities of the six most abundant
lichen species measured on control plots. C)
Radial growth rate of the six most abundant
species measured on disturbed plots.
The Intermediate Disturbance Hypothesis (IDH) has
been frequently invoked over the last 30 years
for its simple prediction that diversity can be
explained as a result of a tradeoff between
colonization and competition at different levels
of disturbance. Despite many documented
patterns of species diversity that appear
consistent with the IDH, few studies actually
test the presumed underlying tradeoff.  
Figure Legend asp Aspicilia spp. rd
Rhizocarpon disporum xl Xanthoparmelia
lineola lm Lecanora muralis pl Unidentified
lichen ls Lecidella stigmataea
A
Kruskal-Wallis chi-squared 3.117, df 5,
p-value 0.682
A preliminary analysis in 1978 of rock dwelling
lichen communities revealed a unimodal
relationship between rates of disturbance
(shearing of the rock face) and diversity of
lichen species (Fig 1). At that time, permanent
plots were established that were followed until
present to test the prediction of a trade off
between colonization ability and competitive
ability as well as other dynamics that could
affect lichen species diversity.
1990
1997
1992
B
Kruskal-Wallis chi-squared 11.010, df 5,
p-value 0.0512
Figure 1. Each point represents a 1 m2 plot in
the Gila National Forest, NM censused for lichen
species in 1978.
2009
2007
2004
Methods
Ten paired plots were established on horizontal
rock faces near the Ben Lily Memorial in the Gila
National Forest NM in 1978. All 10 cm x 10 cm
plots were similar in slope, elevation (2100 m),
and light exposure. One of each pair of plots
was randomly selected, and the entire surface of
the rock face removed with a chisel to simulate
natural disturbance. Beginning in 1978, all
control and disturbed plots were photographed
approximately every two years, resulting in 18
photographs of each of the 20 plots over a
32-year period, allowing us to document lichen
recolonization and competition. The colonization
rate for each lichen species was determined by
counting the number of new colonies on the
disturbed plots over the 32 years. Competitive
abilities were determined by placing a 100 point
grid on photographs of the control plots and
recording the lichen species at each point for
every year photographed. We quantified
competitive ability, Ci, of a lichen species i as
    where Wij is the number of instances a
grid point transitioned from species j to species
i, and Lij is the number of instances a grid
point transitioned from species i to species j,
for the total number of species, n.   To further
explore the community dynamics we also measured
percent cover and average growth rate for each
species. The lichen species at each grid point
in the control plots was recorded, summed over
all plots, and then averaged over each year as a
measure for percent cover. Growth rates were
determined by picking a random individual of each
species on each disturbed plot and quantifying
radial growth over 10 2 years.    Kruskal Wallis
tests were performed to assess significant
differences among species for the response
variables measured (colonization rate,
competitive ability, growth rate, and abundance).
Spearmans rank correlation tests were performed
across plots and across species to assess
relationships between response variables.
C
Kruskal-Wallis chi-squared 14.995, df 5,
p-value 0.0104
Conclusions
We did not find evidence for a simple trade-off
between competitive ability and colonization
ability. Therefore, our results do not support
the mechanisms theorized to contribute to the
patterns described by the IDH.  However, if we
consider a broader suite of traits, we can
explain the majority of the patterns of relative
abundance.  For example, the most abundant
species overall are Aspicilia spp. and
Xanthoparmelia lineola.  Aspicilia has the
highest colonization rate and competitive
ability, while Xanthoparmelia also has a high
competitive ability and the highest growth rate.
Most of the other species have either low
colonization rates or very slow growth rates,
which may explain their low percent cover. This
is not always the case, however (see Lecanora
muralis). A final consideration is the unique
natural history of lichens.  We expected, for
example, that X. lineola might be the best
competitor as its foliose growth form would
overgrow and shade crustose species such as
Aspicilia.  Instead, we observed that Aspicilia
often persisted underneath the X. lineola,
re-appearing when X. lineola flaked off in later
years.  Results of our study suggest that we need
to reconsider how species are successful, beyond
simple colonization and competition traits. .
Figure 3. Percent cover of the six most abundant
species over all control plots averaged over
years.
Kruskal-Wallis chi-squared 80.498, df 5,
p-value lt0.0001
Table 1. Results of Spearman Rho tests for
correlations between colonization, competition,
growth, and abundance among the lichen species.
Table gives Spearman Rho values none are
significant (P gt 0.05).
Acknowledgements
This work was initiated by TEM in 1978 with the
encouragement of Diane Davidson and Jim Brown.
Patricia, Rick, and Bill Miller have helped with
data collection.