|
Defining edge gradients using plant species composition in oak-hickory forest
patches.
Elise M. Tulloss and Scott J. Meiners
Department of Biological Sciences, Eastern Illinois University
Introduction
Forest
conservation is greatly enhanced with an increased understanding of the
transitional nature of edges.
Most forests in east-central Illinois exist as small patches, making the
region an ideal location for studying edge effects.
A quantitative assessment and predictive model of the penetration of edges
into interior forests using plant composition are needed.
Objectives
1) Identify
mechanisms and processes that determine gradients of plant composition and
2) develop a predictive model of edge effects using plant species
composition as an indicator of edge influence.
Methods
A
large data set was gathered to relate species composition to several
environmental and community composition variables.
Data were analyzed in several multiple regressions to find the most important
variables controlling edge responses.
Non-Metric Multidimensional Scaling (NMDS) ordination was performed to determine
if there were compositional changes across the forest edge.
Results
The study
system had a sharply defined edge gradient, with vegetation cover highest at the
edge and decreasing dramatically 15-30 meters into the forest (Figure 1). Total
percent understory cover of plots was the best indicator of edge response for
any of the models (Table 1).
Percent canopy cover, distance from the edge and plot slope were the most
important controllers of edge influence for most responses. Edge orientation had
no effect on response of the community (Figure 2) (P=.961). NMDS produced
3 informative axes that explained 41% of total variation in composition.
Axis 1 and 3 were more strongly correlated with distance from the forest edge
(Figure 3).
|
 |
Figure 1. Percent total cover, proportion of exotic cover,
species richness and proportion of grass coverage based on distance
from the forest edge. |
|
|
|
|
 |
Table 1. Multiple regression equations for community
variables (responses) and environmental variables (predictors). |
|
|
|
|
 |
Figure 2. Total percent cover based on distance from
the forest edge. |
|
|
|
|
 |
Figure 3. Non-Metric Multidimensional Scaling axes 1
and 3. |
|
|
|
Discussion
Canopy cover controlled light intensity to the forest floor.
More open canopies
allowed more light to penetrate and these plots experienced greater understory plant growth.
The distance from the edge of the forest reflected the plot landscape position.
Plot
slopes reflected the control of microtopography on plant growth where steeper plots had less total cover than flatter plots regardless of distance
from the forest edge and understory growth in steep plots was severely limited by light and also leaf
litter from the forest canopy. The lack of effect of forest edge orientation on
the plant community was contrary to most studies of
edge effects in forest-field systems.
Ordination showed Axis 1 is mostly a measure related to forest edges, Axis 2
appeared to explain variation related to different types of edges than Axis 1
and Axis 3 was mostly related to forest interiors
Conclusions
This model serves as a useful tool for predicting the effect of edges on plant
composition and can be calibrated to work for other types of systems. Microtopography can be an important controller of edge effects, which has been
previously overlooked in studies of edge effects
Acknowledgements
BreAnne Nott and
William Stewart.
Tiffany Botany Grant and the
Illinois Native Plant Society.
|
