Journal of Vegetation Science && (2011)
Initial species pattern affects invasion resistance in
experimental grassland plots
Kathryn A. Yurkonis, Brian J. Wilsey & Kirk A. Moloney
Keywords
Andropogon gerardii; Conspecific aggregation;
Monarda fistulosa; Ratibida pinnata;
Schizachyrium scoparium; Spatial pattern;
Tallgrass prairie
Nomenclature
Mohlenbrock (2002)
Received 28 December 2010
Accepted 5 July 2011
Co-ordinating Editor: Amy Symstad
Yurkonis, K.A. (corresponding author,
Kathryn.Yurkonis@email.und.edu): Biology
Department, University of North Dakota, 10
Cornell St., Stop 9091, Grand Forks, ND 58202,
USA
Wilsey, B.J. (bwilsey@iastate.edu) &
Moloney, K.A. (kmoloney@iastate.edu):
Department of Ecology, Evolution &
Organismal Biology, Iowa State University,
253 Bessey Hall, Ames, IA 50011, USA
Abstract
Question: Does the spatial patterning of plant species affect plant community
dynamics in ways that are independent of effects attributable to species richness
or abundances? Does the initial species pattern affect subsequent diversity and
invasion in a perennial grassland system?
Location: Field experiment, Iowa State University Horticultural Research Station, Ames, Iowa, USA.
Methods: Experimental plots (4 m2) were planted with seedlings of four grassland species arranged into increasingly larger groups (patches) of conspecific
individuals while controlling plot-scale richness and evenness. Available light
(photosynthetically active radiation; PAR) at the soil surface was measured each
month for three growing seasons after planting. Species’ relative abundances
were quantified via point-intercept sampling within each plot at the end of each
growing season.
Results: In the three growing seasons after planting, planted species richness
and evenness (plot scale) did not vary among plots planted with different species
patterns. However, early in the second growing season more light reached the
soil surface in plots with initially larger conspecific patches. Invader abundance
was also consistently higher in plots planted with initially larger conspecific
patches.
Conclusions: Our findings support the hypothesis that invasion resistance
increases as communities become more heterogeneous at fine scales and suggest
that within-plot heterogeneity should be considered as an additional factor
when assessing invasion resistance in perennial plant communities.
Introduction
Communities differ from one another in the number (species richness) and abundances (evenness) of constituent
species and both of these attributes affect aspects of diversity maintenance and local invasion (reviewed by Hooper
et al. 2005). Communities may also differ in the ways in
which conspecific and heterospecific individuals are
arranged with respect to one another (Thorhallsdottir
1990; Herben et al. 1993; Purves & Law 2002). Such species patterning may be driven by exogenous (i.e. resource
heterogeneity) or endogenous (i.e. localized dispersal and
competition) factors (Bolker & Pacala 1997). When species
pattern varies among communities established in an
otherwise consistent resource environment, this pattern
may independently affect diversity maintenance and inva-
sibility (Tilman & Kareiva 1997; Murrell et al. 2001; De
Boeck et al. 2006). Although empirical studies have
shown that varying endogenous species patterns affects
the outcome of competitive interactions among neighbouring individuals in annual systems (Norris et al. 2001;
Stoll & Prati 2001; Monzeglio & Stoll 2005) and coral
assemblages (Idjadi & Karlson 2007; Hart & Marshall
2009), the effects of altering the spatial pattern of resident
species have not been empirically tested in species-rich
perennial grasslands, where many studies of the controls
affecting diversity maintenance and invasibility have
occurred.
Varying fine-scale (within-plot) species patterns while
maintaining the same coarser-scale (plot) species richness
and evenness may affect plot-scale responses via two
pathways. First, varying initial species patterns may affect
Journal of Vegetation Science
Doi: 10.1111/j.1654-1103.2011.01331.x © 2011 International Association for Vegetation Science
1
Effects of species pattern on grassland invasion
K.A. Yurkonis et al.
subsequent plot-scale species composition and diversity by
altering intraspecific interactions among establishing seedlings. Establishing seedlings typically interact with one
another over small distances (often at centimetre scales)
and are, thus, most strongly influenced by seedlings in
their immediate vicinity (Murrell & Law 2003; Vogt et al.
2010). Altering species patterns by increasing the size of
patches of conspecific seedlings while maintaining species
abundances at the plot scale will increase the frequency of
intraspecific interactions among neighbours. If the establishing species differ in their inter- and intraspecific competitive abilities (in magnitude or distance over which
interactions occur), then increasing the frequency at
which individuals experience intraspecific interactions
could affect the plot-scale performance of a focal species
(Murrell & Law 2003). At least initially, large conspecific
patches may prevent, or at least delay, competitive exclusion of species more strongly affected by interspecific
competition than intraspecific competition (weak competitors), and result in reduced abundances of species more
strongly affected by intraspecific competition than interspecific competition (strong competitors) (Stoll & Prati
2001; Idjadi & Karlson 2007; Hart & Marshall 2009). However, it is important to note that initial interactions among
seedlings are later affected as mature plants disperse
locally, interact at different distances (neighbourhood size)
(Bolker & Pacala 1999; Law & Dieckmann 2000; Bolker
et al. 2003; Racz & Karsai 2006; Murrell 2010), and interact with the exogenous environment. Consequently, it is
unclear if effects of species patterning during seedling
establishment would persist as communities became more
established.
The effects of initial species pattern on subsequent species composition have been empirically tested in a few
cases. Stoll & Prati (2001) manipulated the spatial patterning of annual species and found that arranging species into
large conspecific patches facilitates persistence of weaker
competitors. These findings were further supported by
studies of the effects of species patterning in tomato and
barnyard grass (Norris et al. 2001), and most recently for
corals (Idjadi & Karlson 2007; Hart & Marshall 2009). Hart
& Marshall (2009) manipulated the patterns of four coral
species and found that some species came to occupy more
area when conspecific individuals were grouped into four
large patches than when they were randomly arranged.
While conducted over short time periods, these experimental results generally support the hypothesis that when
species differ in their relative intra- and interspecific competitive abilities, altering fine-scale species patterning
while maintaining coarser-scale richness and evenness
may affect diversity in the long term. Despite a long-standing interest in how pattern influences process in perennial
systems, this hypothesis has not been tested in species-rich,
2
perennial plant communities (Bolker & Pacala 1999; Bolker et al. 2003).
Varying fine-scale species patterns may also affect invasion at the plot scale (Bergelson 1990b; Olsen et al. 2005;
De Boeck et al. 2006). Establishment in an open site is
generally determined by the density and identity of neighbours around the site (Fowler 1988; Bergelson et al. 1993;
Milbau et al. 2007) and is typically higher in sites surrounded by individuals of a few species than in sites surrounded by individuals of several different species because
resources in low-richness sites are not as completely consumed (Grubb 1977; Naeem et al. 2000; Kennedy et al.
2002). Altering species patterns within a plot may affect
the availability of such low-richness sites for invasion. Plots
with species arranged into large conspecific patches contain more low-richness sites and, therefore, may be less
resistant to invasion than similar plots containing smaller
conspecific patches.
In grasslands, light availability often affects invasion success, and establishment is more likely to be successful in
locations where more light reaches the soil surface (Milbau
et al. 2005; Losure et al. 2007). Thus, light availability at
the soil surface may be used as a proxy for understanding
potential establishment sites for invaders. Because light
use should be lower in low-richness than high-richness
sites (Spehn et al. 2000; De Boeck et al. 2006), mean light
use at the plot scale should be lower in plots with fewer,
larger conspecific patches than in plots with more, smaller
conspecific patches. If invaders are responding to the light
environment, lower mean light use should be correlated
with higher subsequent invader abundance.
In addition to affecting the resource environment for
invasion, species patterns may also affect how an invader
spreads through an area once established (Bergelson et al.
1993; With 2002). If similar unoccupied sites are located
near the establishment site, as would occur when species
occur in large conspecific patches, an invader may more
quickly colonize those sites than if similar unoccupied sites
were further away, as would occur when species occur in
small, conspecific patches. Thus, invaders may be more
abundant when species occur in large conspecific patches
than if the resident species occurred in smaller conspecific
patches because plant arrangement facilitates dispersal of
established invaders.
Only a few studies have considered effects of species
patterns on invasion. Bergelson (1990a) manipulated
arrangements of Poa annua and found invasion by two
annual weeds was higher when seeds were planted in large
conspecific patches than when seeds were randomly
arranged. This effect was due to inhibition of invader
establishment by litter. Olsen et al. (2005) found that crop
planting pattern affected weed recruitment, such that
weeds were less abundant when wheat was planted in a
Journal of Vegetation Science
Doi: 10.1111/j.1654-1103.2011.01331.x © 2011 International Association for Vegetation Science
Effects of species pattern on grassland invasion
K.A. Yurkonis et al.
uniform pattern versus in rows. However, we are not
aware of any studies that have manipulated pattern in
more species-rich communities to test this effect on coarser-scale invasion.
In this study we manipulated species patterns while
maintaining plot-scale species richness and evenness to
test the effects of altering endogenous species patterns on
evenness and invasion. We test the hypotheses that: (1)
evenness; (2) light use ;and (3) invader abundance will
vary among plots with conspecific individuals arranged
into increasingly fewer and larger patches. Our findings
contribute to our understanding of how endogenous species pattern may affect subsequent community dynamics
and are applicable in grassland systems where communities are established in consistent resource environments
(i.e. experimental studies and grassland restorations).
Methods
In May 2006, we planted 24 plots (2 m 9 2 m with 2-m
spacing) with seedlings of four native, tallgrass prairie species in a completely randomized design. Species were
planted into increasingly larger conspecific patches among
plots, but diversity was kept constant at the plot scale by
planting seedlings of all four species in equal abundances
in each plot. Plots were planted at the Iowa State University Horticulture Research Station (Ames, IA; Mean annual
temperature: 8.8 °C; mean annual precipitation: 837 mm).
Vegetation and soils were homogenous across the site prior
to planting. The site was originally dominated by Bromus
inermis Leyss., established on Clarion fine loam soil derived
from glacial till. The area was treated with glyphosate herbicide and disked prior to planting to further homogenize
the soil environment prior to planting.
Seedlings of two C4 bunch grasses [Andropogon gerardii
Vitman and Schizachyrium scoparium (Michx.) Nash] and
two forbs [Ratibida pinnata (Vent.) Barnh. and Monarda
fistulosa L.] that are common in local remnant grasslands
(Martin et al. 2005) were grown in a greenhouse to ca.
equal biomass in a 3:1 sterilized soil:sand mix. A. gerardii
did not grow as well as the other species and, as a result,
seedling biomass differed among species at planting
(F3,20 = 4.72, P = 0.0119). A. gerardii was smaller than
M. fistulosa (0.038 ± 0.0007 g vs 0.117 ± 0.029 g, F1,20 =
6.49, P = 0.0192) and S. scoparium (0.151 ± 0.030 g,
F1,20 = 13.09, P = 0.0017). Seedlings were planted into
bare soil in a grid (8 9 8 plants m 2) with equal spacing
among seedlings (12.5 cm). The seedling density (64
plants m 2) mimicked typical plant densities in local remnant prairies (Losure et al. 2007).
Seedlings were planted following a species array generated for each plot with the program QRULE (Gardner &
Urban 2007). QRULE is a map-generating program that
creates a grid of items based on the desired grid size, the
desired proportion of each item in the grid and the input
H-value, a parameter that affects the likelihood that
adjoining grid locations are assigned to the same item.
H varies from 0 to 1 (Gardner 1999) and arrays created
with a lower H-value are less likely to have adjoining grid
locations assigned to the same item than those created
with a higher H-value. Due to the way in which arrays are
generated, species arrays created with the same H-value
vary from one another. Species arrays were produced by
asking n to create a 16 9 16 array of four species in equal
proportions and with an H-value of 0.0 (n = 8), 0.5 (n = 8)
or 1.0 (n = 8). This approach generated a series of arrays
where conspecific individuals were arranged into increasingly larger groups (patches) (Fig. 1). For our purposes, a
patch was defined as a group of individuals of the same
species occupying neighbouring locations in the planting
grid, where neighbouring individuals could be directly
adjacent to or on a diagonal to one another (Turner et al.
2001) (Fig. 1). Due to the way in which species were
assigned to locations in the arrays, A. gerardii individuals
were more likely to be surrounded by M. fistulosa individuals
than S. scoparium individuals (Type of neighbour, F2,63 =
800.46, P < 0.01). This difference was more pronounced in
arrays with larger conspecific patches (H = 1.0) than in
arrays with smaller conspecific patches (H = 0.0) (Type of
neighbour 9 H-value, F6,63 = 81.72, P < 0.01), as individuals in large conspecific patch arrays were most likely to be
surrounded by conspecific neighbours.
Seedlings were watered upon planting and seedlings
that died were replanted in early Jun 2006. During the first
growing season, plots were weeded to ensure seedling
establishment. In subsequent years, all volunteers were
allowed to persist to test hypotheses about invasion resistance.
In late summer 2007, 2008 and 2009, species abundances were quantified with point-intercept sampling
(a)
(b)
(c)
Fig. 1. Representative planting arrays for plots planted with four common
grassland species (each shaded differently) arranged into increasingly
larger patches across all species. Each pixel represents one individual (256
plot-1) and a patch consists of a group of similarly shaded adjoining pixels.
Planting arrays were created with one of three H-values in the program
QRULE. (a) H = 0.0; area-weighted mean patch size = 0.16 m2, (b)
H = 0.50; area-weighted mean patch size = 0.40 m2, and (c) H = 1.00;
area-weighted mean patch size = 0.68 m2.
Journal of Vegetation Science
Doi: 10.1111/j.1654-1103.2011.01331.x © 2011 International Association for Vegetation Science
3
Effects of species pattern on grassland invasion
K.A. Yurkonis et al.
(Jonasson 1988) over the entire plot. In 2007 and 2008, a
1-m2 sample frame (1.5-m tall) containing a mobile crossbar with holes drilled at 10-cm intervals was placed over
each quadrant of the plot. The frame edges were marked
at 10-cm intervals so that the crossbar could be positioned
to create a grid of 81 (9 9 9) potential sample points
spaced 10-cm apart. Metal pins were dropped vertically
through the vegetation at every other point in this grid
(40 pins m 2). The number of times each species i was
touched by a pin was recorded. Relative abundance of
species i was calculated by dividing the total touches for
species i in a plot (summed across all quadrants) by the
total touches in the plot for each sample year. These data
were used to calculate evenness [(1/D)/S] of the planted
species at the plot scale, where D = ∑pi2, pi = relative
abundance of species i, and S = species richness (Wilsey
et al. 2005). At the end of the 2008 growing season,
above-ground biomass and litter were removed from each
plot in a separate study to assess the utility of using pointintercept sampling to measure biomass-based diversity in
these plots. Because it was impossible to recreate the pattern of ramets pre-harvest, harvested biomass was not
returned to the plots. In 2009, the same sampling method
was used, but the number of sample points was reduced
to 20 random locations in each quadrant to accelerate
sampling. An analysis of the 2008 data indicated that
reducing the number of sample points to 20 would not
affect estimates of plot-scale diversity and invasion. The
difference between evenness and invader abundance calculated from a sub-sample of the points in 2008 (20 randomly selected from each quadrant) and the full sample
sampling effort was not significantly different from zero
(Evenness: t = 0.4045, P > 0.05; Invader abundance:
t = 0.2341, P > 0.05).
To test for effects of species pattern on the light environment, we calculated the proportion of available photosynthetically active radiation (PAR) at the soil surface within
each plot. In 2007 and 2008, above- and below-canopy
(soil surface) PAR was measured each month (May to
Aug) at midday (10:00–14:00 CST). Above- and belowcanopy PAR was measured in four locations in each plot
at each sample date. For each measurement, a Decagon
AccuPAR LP-80 Ceptometer (Pullman, WA, USA) was
inserted along a diagonal line oriented toward the centre
of each quadrant of a plot and a Li-Cor external point sensor (Lincoln, NA, USA) was held above the vegetation. The
Ceptometer simultaneously recorded PAR at the external
point sensor and at each of 80 sensors arranged along a 90cm probe. The proportion of available PAR at the soil surface was calculated by dividing the mean soil surface PAR
for each quadrant by the associated above-canopy PAR.
The proportion of PAR at the soil surface for a plot was the
mean of the quadrant values.
4
Data analysis
Species patterning was quantified by calculating the areaweighted mean patch size (∑ Sk2/∑ Sk; hereafter patch size)
across all of the patches in the species array for each plot,
where Sk is the size of the kth patch in m2. This metric
describes the mean area occupied by a group of conspecific
individuals when plots were planted. Computing mean
patch size in this way reduces the effects of small patch
sizes (e.g. containing one or two individuals) on the overall
mean (Turner et al. 2001). Large values indicate that conspecific individuals were arranged in fewer groups covering a larger area at planting than in plots with small values.
Figure 2 shows the distribution of the patch size variable
for arrays created with each H-value and across all arrays
before and after transformation to improve normality.
Note that the transformed patch size variable differed
among the array categories (F2,21 = 22.8, P < 0.001), and
arrays created with different H-values could have similar
mean patch sizes. Plot evenness and invader abundance
were consistently similar among array categories (results
not shown), indicating that the conditions for generating
the patch size gradient did not affect these responses per se.
Fig. 2. Box and whisker plots of the untransformed and natural logtransformed area-weighted mean patch size variable used as a measure of
initial species patterning. Means are indicated by a dashed line and
outliers are indicated with circles. Separate plots are shown for arrays
created with each H value in QRULE. Percentiles (5th and 95th) could not
be calculated for each array category because of the small sample size
(n = 8) for each. The distribution of the transformed variable is not
significantly different from normal.
Journal of Vegetation Science
Doi: 10.1111/j.1654-1103.2011.01331.x © 2011 International Association for Vegetation Science
Effects of species pattern on grassland invasion
K.A. Yurkonis et al.
To test for effects of species patterning on evenness and
invader abundance in 2007, 2008 and 2009, we used a
repeated-measures ANOVA, with patch size at planting as
a continuous effect (proc glm, SAS 9.2, SAS Institute). For
these and subsequent analyses, patch size was natural logtransformed and invader abundances arcsine square-root
transformed to improve normality.
Finally, monthly mean (natural log-transformed) light
(soil surface PAR) was analysed in separate repeated measures ANOVA (SAS 9.2) for 2007 and 2008, with patch size
as a continuous effect. Correlation between monthly soil
surface PAR and subsequent invader abundance was tested
with Pearson correlation with a Bonferroni adjustment for
multiple comparisons.
Results
All four planted species persisted in all of the plots over the
course of the study. Plot-scale evenness at planting (based
on the number of individuals of each species) was similar
among plots in the three H-value categories (F2,21 = 0.25,
P > 0.05; mean = 0.995 ± 0.0009) and did not vary with
initial patch size (R2 = 0.04, F1,22 = 0.89, P > 0.05). At the
neighbourhood scale (based on the eight neighbours surrounding ten randomly sampled individuals from each
plot), evenness was not correlated with initial patch size
(r = 0.15, P > 0.05), but species richness was negatively
correlated (r = 0.83, P < 0.001) and the probability of
having conspecific neighbours was positively correlated
(r = 0.80, P < 0.001) with initial patch size.
Plot-scale evenness at planting was not maintained after
the initial planting (Table 1), and plot-scale evenness differed among subsequent sample years (F2,44 = 4.42,
P < 0.05). The effect of initial patch size on evenness also
varied among years (Year 9 Size: F2,44 = 3.25, P < 0.05).
Plot-scale evenness was marginally lower in plots planted
with larger initial patches in 2007 (R2 = 0.13; F1,22 = 3.34,
P = 0.08), but not in 2008 (R2 = 0.01; F1,22 = 0.29,
P > 0.05) or 2009 (R2 = 0.19; F1,22 = 1.80, P > 0.05).
Invader abundance differed among years (F2,44 = 69.15,
P < 0.001), increasing over time (Table 1). By 2009,
invaders dominated the space in the plots, occupying ca.
half (52 ± 19%) of the locations originally occupied with a
native individual. Invader abundance was positively
related to initial patch size in all years, but the slope of the
relationship was greater in 2009 than in 2007 and 2008
(F2,44 = 5.01, P < 0.05; Fig. 3). The most abundant invaders included Bromus inermis, Securigera varia (L.) Lassen and
Conyza canadensis (L.) Cronquist, all of which were present
at the site prior to this study.
Light environment
Soil surface PAR was similar among months in 2007
(F3,66 = 0.39, P > 0.05), ranging from 3% of PAR in Jun to
8% of PAR in Jul 2007 (Table 1). The relationship between
soil surface PAR and initial patch size differed among
months (Month 9 Size F3,66 = 4.23, P < 0.01). Soil surface PAR was positively (marginally) related to initial patch
size in May, positively related to initial patch size in Jun,
negatively (marginally) related to initial patch size in Jul
Table 1 Mean plot-scale responses ±1 SE across all plots in an experiment
assessing effects of altering species patterning on plot-scale responses.
Light is given as proportion of photosynthetically active radiation at the soil
surface.
Parameter
2007
2008
2009
Evenness
Invader relative
Abundance
Light
May
June
July
August
0.72 ± 0.025
0.81 ± 0.023
0.70 ± 0.023
0.02 ± 0.004
0.21 ± 0.030
0.55 ± 0.053
0.042
0.034
0.081
0.075
±
±
±
±
0.0082
0.0041
0.0037
0.0030
0.178
0.021
0.038
0.044
±
±
±
±
0.0099
0.0028
0.0029
0.0031
–
–
–
–
Fig. 3. Effect of initial patch size on invader relative abundance in 2007
(note different invader abundance scale), 2008, and 2009 from point
intercept sampling of the above-ground biomass in each plot. *P < 0.05.
Journal of Vegetation Science
Doi: 10.1111/j.1654-1103.2011.01331.x © 2011 International Association for Vegetation Science
5
Effects of species pattern on grassland invasion
K.A. Yurkonis et al.
Fig. 4. Relationship between initial patch size and the proportion of PAR
(photosynthetically active radiation) reaching the soil surface in Jun 2007.
*P < 0.05, **P < 0.10.
and unrelated to patch size in Aug (Fig. 4). In 2008, soil
surface PAR differed among months (F3,66 = 16.29,
P < 0.001), ranging from 17% of PAR in May to 2% of
PAR in Jun (Table 1), and was not related to initial patch
size (F1,22 = 0.16, P > 0.05; Month 9 Size F3,66 = 1.48,
P > 0.05). Finally, yearly invader abundances were not
correlated (P > 0.0025 for all correlations) with soil surface
PAR in the same (2007, 2008) or previous growing seasons
(2008, 2009).
Discussion
We used a plot-based field experiment to test the hypothesis that endogenous species patterning in a grassland system affects planted species interactions and non-planted
species invasion. After four growing seasons, at least some
individuals of all of the planted species were still present in
all of the plots. Planted species evenness was not related to
initial species pattern early in the experiment, when
non-planted (invader) species abundances were low, or in
subsequent growing seasons when invaders were more
abundant. In all years, invader abundance was higher in
plots that started with larger conspecific patches. At least
over the short-term, initial species pattern appears to affect
perennial grassland communities under realistic invasion
pressure via effects on invasion resistance, as opposed to
effects on planted species interactions.
We did not find evidence that species patterning at
establishment affects plot-scale evenness in this system.
Several factors could have contributed to this outcome.
6
First, this result could arise if planted species were similar
in their inter- and intraspecific competitive abilities and
the distances over which inter- and intraspecific interactions took place were equivalent (Murrell & Law 2003). At
least in some cases, effects of inter- and intraspecific interactions on seedling establishment in grasslands are similar
(reviewed in Goldberg & Barton 1992; Milbau et al. 2007;
Vogt et al. 2010). It also appears that, for at least some species, inter- and intraspecific interaction distances are
equivalent (Vogt et al. 2010). Second, the effects of initially increased interspecific interactions among establishing seedlings may have been offset by changes in the
distances over which individuals interacted (neighbourhood size) as they grew. At planting, the effect of a single
seedling on others was likely limited to scales on the order
of 5–15 cm (Milbau et al. 2005; Vogt et al. 2010). As seedlings grew, their physical extent and, thus, their neighbourhood size and the number of individuals they
interacted with increased. Because plots were similar in
composition at larger scales, established individuals may
have experienced similar levels of intraspecific interactions
among the established study plots.
Finally, the degree to which individuals interacted with
conspecific neighbours may have also changed as ramets
of the established individuals dispersed into new locations.
Several theoretical studies have demonstrated that endogenous heterogeneity can facilitate species co-existence
when resident species vary in the extent to which they are
affected by inter- and intraspecific competition and their
abilities to disperse (via ramets or seeds) locally (Bolker &
Pacala 1999; Law & Dieckmann 2000; Bolker et al. 2003;
Racz & Karsai 2006; Murrell 2010). In our study, ramets of
the planted individuals established in the spaces among
the planting locations, altering the neighbourhood composition surrounding the planted individuals. This effect was
most striking for M. fistulosa, a highly clonal forb. Such local
dispersal altered the initial species patterns and likely changed the proportion of intraspecific interactions that resident individuals experienced.
It is unclear if, and to what extent, similarities in interand intraspecific competition and changes in the scale and
types of interaction affected species composition in these
communities. We do not have data concerning inter- and
intraspecific competition coefficients and interaction distances for the species in this study, nor do we know how
these interactions changed as the planted seedlings established and dispersed through the study plots. Future studies need to investigate if local dispersal and changes in the
scale of species interactions over time alter inter- and intraspecific interactions established with the initial species
patterns.
Although there was no evenness response, invasion was
consistently related to species patterning at establishment.
Journal of Vegetation Science
Doi: 10.1111/j.1654-1103.2011.01331.x © 2011 International Association for Vegetation Science
Effects of species pattern on grassland invasion
K.A. Yurkonis et al.
Non-planted species were consistently more abundant in
plots with initially large species patches than in plots with
initially small patches. Our results support the hypothesis
that resistance to invasion increases as communities
become more heterogeneous at fine scales (Davies et al.
2005; Melbourne et al. 2007). However, it is unclear what
mechanisms contributed to this response. Invasion was
not related to variation in light environment established
with the treatments. It is possible that the scale for measuring light was inappropriate, but this is not likely as previous
studies found relationships between light and invasion
using the same sampling methodology as in this system
(Losure et al. 2007). Invasion may have been related to
the ease with which invaders dispersed through a plot once
established (Bergelson et al. 1993; With 2002). Bergelson
et al. (1993) assessed effects of gap size and distribution on
invasion and found that a grassland invader spread more
readily when available gaps for establishment were larger
and closer together. In our study, locations bordered by
the same planted species (spaced 12.5-cm apart) were
more closely positioned in plots with larger initial patch
sizes than in plots with smaller patch sizes (Fig. 1). Thus,
invasion may have increased with initial patch size, as
invaders could more readily disperse into similar sites once
established. To determine the mechanisms behind the
effects of species pattern on invasion, future studies need
to investigate how species patterns affect available
resources for invaders and consider how initial species pattern affects where invaders establish and disperse once
established.
In studying effects of species pattern on plot-scale
responses, we must consider what aspects of the species
pattern affect the responses of interest. Our study controlled two (richness and evenness) of the five major independent components of spatial pattern in landscape maps
identified in Riitters et al. (1995). We quantified effects of
species pattern by assessing plot-scale responses to initial
area-weighted mean patch size, which describes a third
independent component of spatial pattern (perimeter-area
measures) identified in Riitters et al. (1995) and Gardner
& Urban (2007). This area-based metric was used to understand the effects of arranging plants into large, conspecific
patches, as often occurs in grassland restorations (Yurkonis
et al. 2010). In the arrays (16 9 16 cells) used in this
study, the patch size variable was correlated with additional measures that are considered independent in larger
arrays in Riitters et al. (1995) (average patch compaction)
and Gardner & Urban (2007) (number of patches). Initial
patch size was correlated with the average number of species surrounding any given individual, the proportion of
conspecific individuals surrounding any given individual,
and with metrics that describe the shape of patches of conspecific individuals (Yurkonis unpubl. data). Although
these metrics (i.e. neighbourhood richness, proportion of
conspecific individuals and patch shape) may describe ecologically meaningful interactions (Chesson & Neuhauser
2002), we were unable to directly test for their effects with
the methods used to generate species arrays in this study.
Studies that specifically manipulate such aspects of spatial
pattern while holding others constant (i.e. altering patch
shape while maintaining size) could provide a more mechanistic understanding of the effects of species pattern on
plot-scale responses.
Our findings also have implications for how we view
the effects of species patterning in managed and experimental communities established in consistent resource
environments. If species are arranged in different ways as a
community establishes (e.g. De Luis et al. 2008), then this
pattern may affect future invasion, but not planted species
interactions, in realistic communities with relatively
homogeneous resources. In this sense, arranging species in
restored communities into fewer, larger patches (Dickson
& Busby 2009) may hinder restoration success because
such arrangement may facilitate establishment of undesired species.
Conclusions
Connecting pattern and process in plant communities is
challenging because in species-rich communities we cannot easily discern the degree to which individuals are
affected by their neighbours. As a few other studies have
done (Stoll & Prati 2001; Idjadi & Karlson 2007; Hart &
Marshall 2009), we approach the relationship between
pattern and process in communities by testing for the overall consequences of altering endogenous species pattern
within a grassland community. In a perennial grassland
system, it appears that initial species pattern may affect
invasion resistance, but not planted species evenness or
richness. Future studies need to separate effects of endogenous species pattern on planted species interactions from
the effects on invasion resistance to assess if these findings
are consistent in communities with less invasion pressure.
Acknowledgements
Many thanks to the Iowa State University Horticulture Station staff who helped to prepare and manage the site. A.
Asche, A. Blong, A. Conner, T. Dickson, T. Moeller, J. Reynolds, J. Richter, C. Swenson, K. Wahl and S. Yurkonis
assisted in planting and sampling the plots. F. I. Isbell provided valuable feedback throughout the project. Amy
Symstad and two anonymous reviewers provided helpful
comments on earlier drafts. This project was partially
funded by the Iowa Department of Transportation Living
Roadway Trust Fund and Prairie Biotic Research Inc.
Journal of Vegetation Science
Doi: 10.1111/j.1654-1103.2011.01331.x © 2011 International Association for Vegetation Science
7
Effects of species pattern on grassland invasion
K.A. Yurkonis et al.
References
Bergelson, J. 1990a. Life after death: site pre-emption by the
remains of Poa annua. Ecology 71: 2157–2165.
Bergelson, J. 1990b. Spatial patterning in plants: opposing
effects of herbivory and competition. Journal of Ecology 78:
937–948.
Bergelson, J., Newman, J.A. & Floresroux, E.M. 1993. Rates
of weed spread in spatially heterogeneous environments.
Ecology 74: 999–1011.
Bolker, B. & Pacala, S.W. 1997. Using moment equations to
understand stochastically driven spatial pattern formation in
ecological systems. Theoretical Population Biology 52: 179–197.
Bolker, B.M. & Pacala, S.W. 1999. Spatial moment equations
for plant competition: understanding spatial strategies and
the advantages of short dispersal. American Naturalist 153:
575–602.
Bolker, B.M., Pacala, S.W. & Neuhauser, C. 2003. Spatial
dynamics in model plant communities: what do we really
know? American Naturalist 162: 135–148.
Chesson, P. & Neuhauser, C. 2002. Intraspecific aggregation and
species coexistence. Trends in Ecology and Evolution 17:
210–211.
Davies, K.F., Chesson, P., Harrison, S., Inouye, B.D., Melbourne,
B.A. & Rice, K.J. 2005. Spatial heterogeneity explains the
scale dependence of the native–exotic diversity relationship.
Ecology 86: 1602–1610.
De Boeck, H.J., Nijs, I., Lemmens, C.M.H.M. & Ceulemans, R.
2006. Underlying effects of spatial aggregation (clumping) in
relationships between plant diversity and resource uptake.
Oikos 113: 269–278.
De Luis, M., Raventós, J., Wiegand, T. & González-Hidalgo, J.C.
2008. Temporal and spatial differentiation in seedling emergence may promote species coexistence in Mediterranean
fire-prone ecosystems. Ecography 31: 620–629.
Dickson, T.L. & Busby, W.H. 2009. Forb species establishment
increases with decreased grass seeding density and with
increased forb seeding density in a Northeast Kansas, USA,
experimental prairie restoration. Restoration Ecology 17: 597–
605.
Fowler, N.L. 1988. What is a safe site – neighbor, litter, germination date, and patch effects. Ecology 69: 947–961.
Gardner, R.H. 1999. RULE: Map generation and a spatial analysis
program. In: Klopatek, J.M. & Gardner, R.H. (eds.) Landscape
ecological analysis: issues and applications. pp. 280–303.
Springer, New York, NY, US.
Gardner, R.H. & Urban, D.L. 2007. Neutral models for testing
landscape hypotheses. Landscape Ecology 22: 15–29.
Goldberg, D.E. & Barton, A.M. 1992. Patterns and consequences
of interspecific competition in natural communities: a
review of field experiments with plants. American Naturalist
139: 771–801.
Grubb, P.J. 1977. The maintenance of species richness in plant
communities: the importance of the regeneration niche. Biological Review 52: 107–145.
8
Hart, S.P. & Marshall, D.J. 2009. Spatial arrangement affects population dynamics and competition independent of community composition. Ecology 90: 1485–1491.
Herben, T., Krahulec, F., Hadincova, V. & Kovarova, M. 1993.
Small-scale spatial dynamics of plant species in a grassland
community over six years. Journal of Vegetation Science 4:
171–178.
Hooper, D.U., Chapin, F.S., Ewel, J.J., Hector, A., Inchausti, P.,
Lavorel, S., Lawton, J.H., Lodge, D.M., Loreau, M., Naeem,
S., Schmid, B., Setala, H., Symstad, A.J., Vandermeer, J. &
Wardle, D.A. 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs 75: 3–35.
Idjadi, J.A. & Karlson, R.H. 2007. Spatial arrangement of competitors influences coexistence of reef-building corals. Ecology
88: 2449–2454.
Jonasson, S. 1988. Evaluation of the point intercept method for
the estimation of plant biomass. Oikos 52: 101–106.
Kennedy, T.A., Naeem, S., Howe, K.M., Knops, J.M.H., Tilman,
D. & Reich, P. 2002. Biodiversity as a barrier to ecological
invasion. Nature 417: 636.
Law, R. & Dieckmann, U. 2000. A dynamical system for neighborhoods in plant communities. Ecology 81: 2137–2148.
Losure, D.A., Wilsey, B.J. & Moloney, K.A. 2007. Evenness–
invasibility relationships differ between two extinction
scenarios in tallgrass prairie. Oikos 116: 87–98.
Martin, L.M., Moloney, K.A. & Wilsey, B.J. 2005. An assessment
of grassland restoration success using species diversity components. Journal of Applied Ecology 42: 327–336.
Melbourne, B.A., Cornell, H.V., Davies, K.F., Dugaw, C.J., Elmendorf, S., Freestone, A.L., Hall, R.J., Harrison, S., Hastings, A., Holland, M., Holyoak, M., Lambrinos, J., Moore, K.
& Yokomizo, H. 2007. Invasion in a heterogeneous world:
resistance, coexistence or hostile takeover? Ecology Letters 10:
77–94.
Milbau, A., Nijs, I., De Raedemaecker, F., Reheul, D. & De Cauwer, B. 2005. Invasion in grassland gaps: the role of neighbourhood richness, light availability and species
complementarity during two successive years. Functional
Ecology 19: 27–37.
Milbau, A., Reheul, D., De Cauwer, B. & Nijs, I. 2007. Factors
determining plant–neighbour interactions on different spatial scales in young species-rich grassland communities.
Ecological Research 22: 242–247.
Mohlenbrock, R.H. 2002. Vascular Flora of Illinois. Southern Illinois University Press, Carbondale, IL, US.
Monzeglio, U. & Stoll, P. 2005. Spatial patterns and species performances in experimental plant communities. Oecologia
145: 619–628.
Murrell, D.J. 2010. When does local spatial structure hinder
competitive coexistence and reverse competitive hierarchies? Ecology 91: 1605–1616.
Murrell, D.J. & Law, R. 2003. Heteromyopia and the spatial
coexistence of similar competitors. Ecology Letters 6: 48–59.
Journal of Vegetation Science
Doi: 10.1111/j.1654-1103.2011.01331.x © 2011 International Association for Vegetation Science
Effects of species pattern on grassland invasion
K.A. Yurkonis et al.
Murrell, D.J., Purves, D.W. & Law, R. 2001. Uniting pattern and
process in plant ecology. Trends in Ecology and Evolution 16:
529–530.
Naeem, S., Knops, J.M.H., Tilman, D., Howe, K.M., Kennedy, T.
& Gale, S. 2000. Plant diversity increases resistance to invasion in the absence of covarying extrinsic factors. Oikos 91:
97–108.
Norris, R.F., Elmore, C.L., Rejmanek, M. & Akey, W.C. 2001.
Spatial arrangement, density, and competition between
barnyardgrass and tomato: I. Crop growth and yield. Weed
Science 49: 61–68.
Olsen, J., Kristensen, L., Weiner, J. & Griepentrog, H.W. 2005.
Increased density and spatial uniformity increase weed suppression by spring wheat. Weed Research 45: 316–321.
Purves, D.W. & Law, R. 2002. Fine-scale spatial structure in a
grassland community: quantifying the plant’s-eye view.
Journal of Ecology 90: 121–129.
Racz, E.V.P. & Karsai, J. 2006. The effect of initial pattern on
competitive exclusion. Community Ecology 7: 23–33.
Riitters, K.H., O’Neill, R.V., Hunsaker, C.T., Wickham, J.D.,
Yankee, D.H., Timmins, S.P., Jones, K.B. & Jackson, B.L.
1995. A factor analysis of landscape pattern and structure
metrics. Landscape Ecology 10: 23–39.
Spehn, E.M., Joshi, J., Schmid, B., Diemer, M. & Korner, C.
2000. Above-ground resource use increases with plant species richness in experimental grassland ecosystems. Functional Ecology 14: 326–337.
Stoll, P. & Prati, D. 2001. Intraspecific aggregation alters competitive interactions in experimental plant communities. Ecology
82: 319–327.
Thorhallsdottir, T.E. 1990. The dynamics of a grassland community: a simultaneous investigation of spatial and temporal
heterogeneity at various scales. Journal of Ecology 78: 884–
908.
Tilman, D. & Kareiva, P.M. (eds.) 1997. Spatial ecology: the role of
space in population dynamics and interspecific interactions. Princeton University Press, Princeton, NJ, US.
Turner, M.G., Gardner, R.H. & O’Neill, R.V. 2001. Landscape ecology in theory and practice: pattern and process. Springer, New
York, NY, US.
Vogt, D.R., Murrell David, J. & Stoll, P. 2010. Testing spatial theories of plant coexistence: no consistent differences in intraand interspecific interaction distances. American Naturalist
175: 73–84.
Wilsey, B.J., Chalcraft, D.R., Bowles, C.M. & Willig, M. 2005.
Relationships among indices suggest that richness is an
incomplete surrogate for grassland biodiversity. Ecology 86:
1178–1184.
With, K.A. 2002. The landscape ecology of invasive spread. Conservation Biology 16: 1192–1203.
Yurkonis, K.A., Wilsey, B.J., Moloney, K.A., Drobney, P. &
Larson, D.L. 2010. Seeding method influences warm-season
grass abundance and distribution but not local diversity in
grassland restoration. Restoration Ecology 18: 344–353.
Journal of Vegetation Science
Doi: 10.1111/j.1654-1103.2011.01331.x © 2011 International Association for Vegetation Science
9