Content uploaded by Luke Flory
Author content
All content in this area was uploaded by Luke Flory on Jan 14, 2019
Content may be subject to copyright.
ORIGINAL PAPER
Propagule quality mediates invasive plant establishment
James A. Estrada .Chris H. Wilson .
Julienne E. NeSmith .S. Luke Flory
Received: 7 August 2015 / Accepted: 9 May 2016 / Published online: 18 May 2016
ÓSpringer International Publishing Switzerland 2016
Abstract Propagule pressure is commonly consid-
ered a primary driver of invasive plant establishment
and spread. However, the physical size or condition
(i.e., quality) of propagules may also affect establish-
ment, particularly under unfavorable habitat condi-
tions such as low light environments. We used an
outdoor mesocosm experiment to test the relative
contribution of propagule size (number of individuals
introduced) and quality (number of rhizome nodes) to
the establishment and performance of the highly
invasive cogongrass (Imperata cylindrica) under
experimental sun and shade treatments. We found
that the introduction of higher quality propagules
(rhizome segments C3 nodes in length) significantly
enhanced establishment across both light treatments,
and increased final tiller count in the sun treatment.
The sun treatment also enhanced rhizome growth, an
effect that could increase spread rates and invasion
success. Thus, while cogongrass is likely to establish
in both sun and shade, introductions of large propagule
sizes or large rhizomes in high light environments
likely poses the greatest threat to native habitats. Our
results demonstrate that propagule quality promoted
both establishment and performance of a highly
invasive grass species and suggest that propagule
quality may play an important but underappreciated
role in the invasion process.
Keywords Cogongrass Propagule pressure
Imperata cylindrica Invasion Mechanism
Introduction
Understanding the mechanisms that underlie non-
native plant invasions is essential to predicting
vulnerable habitats and future invaders. However,
identifying factors that are general predictors of
invasion success has proven difficult because the
establishment of invasive plants is unlikely to be
controlled by a single mechanism, but rather complex
interactions among multiple determinants (Gurevitch
et al. 2011; Inderjit et al. 2005; Lau and Schultheis
2015). Nonetheless, propagule pressure has emerged
as a common driver of plant establishment across a
variety of habitats and species (Lockwood et al. 2009;
Simberloff 2009) and is often positively correlated
with invasion success (Blackburn et al. 2009; Lock-
wood et al. 2009; Simberloff 2009; Williamson 1996).
Electronic supplementary material The online version of
this article (doi:10.1007/s10530-016-1163-9) contains supple-
mentary material, which is available to authorized users.
J. A. Estrada (&)S. L. Flory
Agronomy Department, University of Florida,
Gainesville, FL 32611, USA
e-mail: estradaj@ufl.edu
C. H. Wilson J. E. NeSmith
School of Natural Resources and Environment, University
of Florida, Gainesville, FL 32611, USA
123
Biol Invasions (2016) 18:2325–2332
DOI 10.1007/s10530-016-1163-9
Although propagule pressure has been frequently used
to explain the success or failure of past introductions,
studies examining its ability to predict future invasions
remain rare (reviewed in Lockwood et al. 2005).
While the term ‘‘propagule’’ can be defined as
either a group of individuals or reproductive units
(e.g., 100 seeds) or a single individual (a single seed,
piece of stem, or rhizome), the concept of ‘‘propagule
pressure’’ is well defined. It is comprised of two
components: propagule size (number of individuals
introduced) and propagule number (the rate at which
propagules arrive over time; Lockwood et al. 2005;
Simberloff 2009). Thus, propagule pressure represents
the total number of individuals introduced per unit
time. Moreover, both propagule size and number have
been shown to independently increase the likelihood
of establishment (Lockwood et al. 2005; Simberloff
2009).
The mass or physical condition of a propagule
(hereafter referred to as propagule quality) may also
influence the establishment success of invasive plants
(Ruiz et al. 2000; Williamson 1996). For example,
large seed size (i.e., high quality) has been shown to
enhance seed longevity (Moles et al. 2000), germina-
tion (Eisenhauer and Scheu 2008; Moles and Westoby
2004a), and seedling survival (Moles and Westoby
2004b). In particular, large seeds may be beneficial
under adverse conditions or hazards such as shade,
drought, competition, or herbivory (Flory and Clay
2010; Leishman and Westoby 1994; Lonnberg and
Eriksson 2013; Moles and Westoby 2004a), where
stored energy reserves may help to overcome envi-
ronmental stressors and stochasticity. Seed size is
often greater in introduced than native ranges for
invasive plant species, due to enemy release, post-
introduction evolution, or differences in environmen-
tal conditions (Daws et al. 2007). In turn, such
differences in seed size may provide a competitive
advantage for non-native plants in the invasive range.
In contrast, it has also been reported that small seed
size is related to greater invasiveness in some species
(Rejma
´nek and Richardson 1996; Richardson and
Rejmanek 2004), and that small seed mass is corre-
lated with invasion success at larger spatial scales
(Hamilton et al. 2005). Small seed size may be
advantageous because it is often associated with
higher overall seed production (Greene and Johnson
1994; Primack 1978) and enhanced dispersal (Harper
et al. 1970; Van Wilgen and Siegfried 1986). Whether
large or small propagule mass is more advantageous
may be heavily dependent on both species (reproduc-
tive strategy, germination rates, seedling vigor) and
habitat (environmental conditions, disturbance, dis-
persal barriers) characteristics. However, experimen-
tal studies on the role of propagule quality in the
establishment of invasive species are rare (Lockwood
et al. 2005; Simberloff 2009) and limited primarily to
observational investigations or animal studies, such as
ballast introductions (e.g. Ruiz et al. 2000). Further-
more, the likelihood of introducing high quality
propagules is expected to increase with greater
propagule size and number. Thus, studies comparing
the relative influence of propagule quality, size, and
number across environmental conditions are needed.
The effects of propagule quality may be particularly
relevant for plants with vegetative reproduction,
another trait frequently associated with invasion
success (Kolar and Lodge 2001; Lloret et al. 2005;
Quinn and Holt 2009), because the physical size or
condition of the propagule may vary substantially
among introduction events. It is likely that larger
pieces of rhizomes or stem fragments (higher quality)
would be beneficial for species with vegetative
reproduction because greater mass would mean higher
carbon stores to support initial establishment and
growth. Furthermore, a higher quality propagule may
be especially important in low resource environments,
such as low light, where establishment might rely
heavily on energy stores of the vegetative fragment.
However, only a single study (Quinn and Holt 2009)
has investigated the influence of propagule quality on
the performance of an invasive plant species with
vegetative reproduction. Additionally, while they
reported that propagule mass contributed positively
to invader shoot height and performance, the relation-
ship did not exist at all study sites. Thus, further
research to empirically evaluate the effects of propag-
ule quality on invasive plant establishment is needed.
In this study we used an outdoor mesocosm
experiment to test the relative contribution of propag-
ule size and quality to the establishment success and
performance of cogongrass (Imperata cylindrica, (L.)
P. Beauv., hereafter cogongrass), a highly invasive,
rhizomatous C
4
grass. Cogongrass is native to Asia
and has invaded a wide variety of habitats (e.g., natural
areas, pine plantations, pastures) throughout the
eastern US from Texas to Florida and as far north as
Virginia (USDA, NRCS 2005). It is considered a
2326 J. A. Estrada et al.
123
primary threat to native biodiversity and ecosystem
functions (Brewer 2008; Estrada and Flory 2015;
MacDonald 2004) and is a federally listed noxious
weed (USDA, NRCS 2005). The accidental transport
of cogongrass rhizome fragments (e.g., via fill dirt or
transport on machinery, Willard et al. 1990) is thought
to be a common means of introduction throughout
much of the invasive range. Since both the number and
quality of rhizome fragments likely varies substan-
tially among introduction events, we tested whether
increasing propagule size (number of propagules) and
quality (length/mass of rhizome fragments) would
enhance cogongrass establishment. We were inter-
ested in evaluating a single introduction event, thus
propagule number was not manipulated. As with many
plant species, cogongrass establishment is highly
influenced by light availability (Ayeni and Duke
1985), so we also evaluated whether greater propagule
quality would improve establishment probability in
lower light environments. We predicted that both
greater propagule size and quality would positively
influence cogongrass performance, and that higher
quality propagules would better establish and perform
under low resource, shaded conditions.
Methods
Experimental design
To evaluate the contributions of propagule size and
propagule quality to cogongrass establishment and
performance under sun and shade, we established an
outdoor mesocosm experiment at the Bivens Arm
Research Site (BARS) in Gainesville, FL (29.628489°
N, -89.353370°W). We introduced three propagule
sizes (one, three, and five rhizome fragments) and
three propagule quality treatments (one, three, and five
nodes) in a factorial design (nine total treatment
combinations) nested within two light treatments with
ten replicates (180 mesocosms total). Since the
diameters of cogongrass rhizomes vary, we standard-
ized the treatments by mass, with one, three, and five
node segments weighing 0.17 ±0.02, 0.46 ±.04,
and 0.76 ±0.07 g (mean ±SD) respectively. We
obtained the ranges by calculating a 15 % interval
around the mean mass of each quality treatment.
Rhizome fragments were planted 2–3 cm deep in 3.8L
round, plastic pots filled with a heterogeneous mixture
of screened local topsoil (Florida Green Keepers,
Alachua County, FL). We then arranged the pots into
20 plots and randomly assigned each plot to either full
sun (n =10) or 60 % shade cloth (n =10). The
experiment was conducted for a period of 12 weeks
(August through November 2013). All pots were
watered daily and a 50 % concentration of 24-8-16
water-soluble fertilizer (Scotts Miracle-Gro Products,
Inc. Marysville, OH) was applied during the 7th week
of the experiment. Rhizome fragments were collected
from an existing cogongrass population at BARS.
Data collection
We recorded the total number of tillers and height of
the tallest tiller in each mesocosm twice per week for
the first 10 weeks of the experiment and once per week
for the final 2 weeks. We scored mesocosms as having
‘‘established’’ cogongrass if live shoots were present
in the mesocosm at the conclusion of the experiment.
All biomass was harvested in November 2013 and
aboveground and belowground portions were sepa-
rated. During the harvest we located and removed the
originally introduced rhizome segments to ensure that
final belowground values only reflected biomass
produced during the course of the experiment. Root
and rhizome materials were washed to removed soil,
and all plant materials were dried at 60 °C to constant
mass and weighed.
Statistical analysis
We analyzed four response variables using general-
ized linear mixed effects models: establishment
(presence of tillers at the conclusion of the experi-
ment), final tiller count, aboveground biomass, and
belowground biomass. For all models, we included
categorical predictor variables for fixed effects of light
treatment (sun/shade), propagule size (one, three, or
five propagule fragments), and propagule quality (one,
three, or five nodes), alongside plot as a random effect
to account for the blocking of mesocosms within light
treatments in our experimental design. For statistical
inference, we examined the posterior distributions of
the fixed effects predictors to determine if the 95 %
confidence intervals excluded zero (analogous to a
Wald-Z or a likelihood ratio test, Bolker et al. 2009).
Propagule quality mediates invasive plant establishment 2327
123
For categorical predictors, these parameter estimates
represent contrasts to an estimated intercept term that
serves as the baseline (i.e., reference) level. These
results are displayed graphically in coefficient plots to
facilitate ready interpretation and comparison of both
effect size and uncertainty across predictors.
We modeled establishment as a Bernoulli random
variable (0/1) related to our predictor variables via the
logit-link function. The subsequent analyses (tiller
count, biomass) were conducted using data only from
mesocosms where cogongrass had emerged after
4 weeks (i.e., removing the zeroes due to establish-
ment failure). The above and belowground biomass
data were analyzed with a normal distribution. The
tiller production data were assigned a Poisson distri-
bution, including a random effect term for overdis-
persion (Gelman and Hill 2006) and given the standard
log-link.
We were specifically interested in the effect of
propagule quality on tiller counts since we observed a
notable advantage of three and five node rhizome
segments over single node segments in the sun
treatment. Therefore we created a set of six test
statistics (T
test
) that contrasted predictions of expected
tiller count under all six combinations of propagule
size and light treatment among mesocosms with three
or five node segments versus mesocosms with the
same propagule size and light treatment but with
single node segments. Under the Poisson model, these
expectations are simply the exponentiated sum of our
fitted fixed effects parameters (Gelman and Hill 2006).
We computed the probability that the T
tests
were
positive, indicating an advantage for a higher quality
propagule, by sampling their posterior distributions
(i.e., computing proportion of draws above zero). We
graph the mean and the 50 and 95 % credible intervals
for the T
test
and highlight the results where the 95 %
intervals exclude zero.
All models were estimated via Gibbs-sampling
Markov Chain Monte Carlo methods, using 5 chains of
12,000 iterations each, discarding the first 2000
iterations, and thinning by a factor of 50, leading to
1000 draws from the posterior distribution for each
parameter. Convergence was assessed via the Gelman-
Rubin diagnostic (R-hat \1.1, Gelman et al. 2013)
and by visually checking for well-blended chains. All
models were programmed in R (v. 3.1.0, R Core
Development Team) and JAGS (v.3.4.0), linked via
the ‘‘R2jags’’ package.
Results
Establishment
We found strong evidence for the role of both propagule
size and quality in predicting establishment (presence
of tillers after 12 weeks). The main effects indicators
for three and five propagule treatments were significant
and positive (mean: 2.59, 95 % CI 018–5.343; 4.79,
2.003–7.995, respectively), while the main effect
indicator for the three node quality treatment was also
significant and positive (2.54, 1.359–5.343, Fig. 1a).
Neither light treatment nor the five node treatment were
significant, nor were any of the higher-order
interactions.
Tiller count
Tiller count was most strongly driven by propagule
size, with the five propagule coefficient significant and
positive (1.25, 0.462–2.093, Fig. 1b). Our test statis-
tics, which contrasted tiller count predictions for three
node as opposed to single node rhizome segments
under all combinations of propagule size and light
treatments, predicted an advantage (median) for a
higher quality propagule in sun treatments when
propagule size was above one (Fig. A1). Unlike the
emergence model, the light treatment was significant
and negative (-1.382, -2.694 to -0.101) indicating
that tiller count was suppressed by shading. This result
can also be seen in the raw data on tiller counts, which
indicated a 200 % increase in tillers for plants in the
sun compared to shade treatment across all propagule
size and quality measures (Fig. A2).
Biomass and plant height
Aboveground biomass was determined almost exclu-
sively by propagule size, with the main effect indictor
for the five propagule treatment being the only positive
and significant coefficient (1.71, 0.717–3.171, Fig. 2a).
While the three node treatment suggested a net positive
result, the considerable overlap with zero precludes
making a solid statistical inference. Neither the light
environment nor propagule quality had a discernible
impact on final aboveground biomass. There was,
however, a significant difference in plant height, with
those in full sun (mean ±SD: 0.54 ±0.63 cm) being
2328 J. A. Estrada et al.
123
shorter than those in shaded treatments
(1.12 ±1.13 cm) (t =13.75, p\0.0001).
Similar to the aboveground biomass model, below-
ground biomass was also driven by propagule size, with
no significant influence from the propagule quality
treatments (Fig. 2b). Both the three and five propagule
treatments showed positive and significant main effect
indicators (1.30, -0.011 to 2.516; 1.76, 0.631–2.929,
respectively) demonstrating that introductions of three
or five rhizome fragments produced significantly more
belowground biomass than single rhizome introduc-
tions. Belowground biomass was also influenced by
light environment, with reduced rhizome production
under the shade treatment (-1.312, -2.890 to 0.154).
Total belowground biomass for plants in the sun
(mean ±SD: 1.11 ±0.13 g) was on average twice
as high as for those in the shaded treatment
(0.54 ±0.07 g).
Fig. 1 Coefficient plots for establishment (a) and tiller count
(b). Each coefficient represents positive or negative treatment
effects as compared to a reference level of one propagule with
one node in the sun. We consider coefficients with confidence
intervals (95 %) excluding zero to be significantly positive or
negative, with those above zero being positive and below being
negative
Fig. 2 Coefficient plots for aboveground (a) and belowground
(b) biomass. Each coefficient represents positive or negative
treatment effects as compared to a reference level of one
propagule with one node in the sun. We consider coefficients
with confidence intervals (95 %) excluding zero to be signif-
icantly positive or negative, with those above zero being
positive and below being negative
Propagule quality mediates invasive plant establishment 2329
123
Discussion
Here we report on the first experimental evaluation of
the relative influence of propagule pressure and
propagule quality on the establishment and perfor-
mance of an invasive plant species. Our finding that
propagule size was the most influential factor across
all treatments supports previous studies indicating that
propagule pressure is a primary driver of non-native
plant establishment (reviewed in Lockwood et al.
2005; Simberloff 2009). We also showed that propag-
ule quality significantly enhanced both establishment
and performance. However, in contrast to our predic-
tions, greater propagule quality did not increase
establishment or performance in the shade compared
to sun treatment. Thus, cogongrass is likely to
establish in habitats with a wide range of light
conditions, but performance will be greatest for large
propagule sizes and high quality propagules under full
sun. More generally, our data suggest that propagule
quality may play an important but underappreciated
role in the invasion process.
Although we found that higher propagule quality
(Cthree node rhizome fragment) enhanced establish-
ment, there was no additional benefit of five compared
to three nodes. This result was unexpected because the
presence of additional nodes (i.e., potential growth
sites) should presumably increase the likelihood of
establishment. Additionally, it was assumed the larger
five node rhizome segments would contain more
stored energy (Decruyenaere and Holt 2001; Dong and
Pierdominici 1995; Quinn and Holt 2008), thereby
promoting establishment. The advantage of the three
node segments was driven by extremely high estab-
lishment ([98 %) across both light treatments when
more than one propagule was introduced, indicating
that rhizome fragments of this quality present a
considerable invasion risk. Yet, nearly 50 % of single
rhizome fragments with one node established. It is
possible that the incorporation of additional stressors
(e.g., drought, competition) may have reduced the
viability of the single node fragments, but our results
suggest that even a very small propagule may result in
establishment and potentially invasions.
While our data are in general agreement with
previous work on propagule quality, our findings
provide two novel insights. First, propagule quality
had a larger influence over establishment success than
the light environment. This finding is important because
light availability is a well-known driver of plant
performance and can dictate establishment success for
many non-native species (Canham et al. 1990; Val-
ladares 2003). However, there is evidence suggesting
that established cogongrass populations might tolerate
reduced light environments. For example, studies
conducted by Patterson (1980) suggest that cogongrass
is able to persist in up to 50 % shade and rapidly adapt
to changes in light levels by altering specific leaf area
and leaf area ratio. Gaffney (1996) and Ramsey et al.
(2003) also found cogongrass to have a low light
compensation point (32–35 lmol m
-2
s
-1
), indicating
that it could thrive as an understory species. Establish-
ment in shaded habitats has been reported for other
invasive grasses, including giant cane (Arundo donax,
Quinn and Holt 2009) and stiltgrass (Microstegium
vimineum, Schramm and Ehrenfeld 2010;Wilsonetal.
2015), but in general, reduced light availability is
thought to inhibit the survival and spread of invasive
grasses (Funk and McDaniel 2010;LohandDaehler
2007). The long-term fate of newly established
cogongrass in shaded habitats has not been evaluated
so it is unclear whether these populations will ulti-
mately spread or die off once energy stores in the
propagule are depleted. Our second important conclu-
sion is that even ‘‘low quality’’ (i.e., one node)
propagules may present a substantial invasion risk
under favorable environmental conditions. While high
light availability did not increase establishment success,
it significantly enhanced tiller production relative to the
shade treatment. Thus, although the introduction of a
single node rhizome segment presents considerably less
risk than larger fragments, those that do survive and
establish are expected to rapidly spread in full sun
habitats.
Light availably has been shown to drive below-
ground production in grasses, often resulting in
rhizome proliferation under high light levels (Dong
and Pierdominici 1995; Wills 1975). Accordingly, our
results show a significant increase in rhizome biomass
in the sun treatment. Accumulation of belowground
biomass in cogongrass may be an important compo-
nent of invasion success since vast rhizome assem-
blages have been linked to its ability to rapidly regrow
and dominate landscapes after disturbances such as
mowing or fire (Lippincott 2000), making control
difficult. Additionally, enhancement of belowground
biomass likely explains the dramatic increase in tiller
counts in the sun treatment since this pattern of growth
2330 J. A. Estrada et al.
123
would also generate more nodes for shoot develop-
ment. We therefore conclude that invasions in habitats
with high light availability are likely to rapidly spread
and complicate management efforts.
Our results demonstrate that propagule size was the
most influential factor in determining establishment
and performance of a highly invasive grass species,
but that propagule quality also plays an important role
in cogongrass invasions. Furthermore, we showed that
propagule quality might be a more influential deter-
minant of establishment than light availability. Given
the relative lack of studies that have examined
propagule quality in the invasion process we urge that
future research efforts be focused on: (1) measuring
natural variations in propagule quality in the field, (2)
experimental introductions of different propagule
qualities across variable habitat characteristics (i.e.,
resource availability, resident species diversity and
density, disturbance level) to help improve predictions
of invasion success, and (3) determining the extent to
which propagule quality mitigates environmental or
demographic stochasticity during introductions. In
conclusion, our study shows that the introduction of
high quality propagules can enhance invasive plant
establishment, even at low propagule sizes, suggesting
that propagule quality is an important component of
predicting invasive plant establishment.
Acknowledgments We thank Deah Lieurance, Christina
Alba, and members of the Flory Lab for helpful discussions
and revisions on earlier versions of the manuscript. Support was
provided in part by the Florida Fish and Wildlife Conservation
Commission.
References
Ayeni AO, Duke WB (1985) The influence of rhizome features
on subsequent regenerative capacity in speargrass (Im-
perata cylindrica (L) Beauv). Agric Ecosyst Environ
13:309–317
Blackburn TM, Cassey P, Lockwood JL (2009) The role of
species traits in the establishment success of exotic birds.
Glob Change Biol 15:2852–2860
Bolker BM, Brooks ME, Clark CJ, Geange SW, Poulsen JR,
Stevens MHH, White J-SS (2009) Generalized linear
mixed models: a practical guide for ecology and evolution.
Trends Ecol Evol 24:127–135
Brewer S (2008) Declines in plant species richness and endemic
plant species in longleaf pine savannas invaded by Im-
perata cylindrica. Biol Invasions 10:1257–1264
Canham CD, Denslow JS, Platt WJ, Runkle JR, Spies TA, White
PS (1990) Light regimes beneath closed canopies and tree-
fall gaps in temperate and tropical forests. Can J For Res
20:620–631
Daws MI, Hall J, Flynn S, Pritchard HW(2007) Do invasive
species have bigger seeds? Evidence from intra- and inter-
specific comparisons. S Afr J Bot 73:138–143
Decruyenaere JG, Holt JS (2001) Seasonality of clonal propa-
gation in giant reed. Weed Sci 49:760–767
Dong M, Pierdominici MG (1995) Morphology and growth of
stolons and rhizomes in three clonal grasses, as affected by
different light supply. Vegetatio 116:25–32
Eisenhauer N, Scheu S (2008) Invasibility of experimental
grassland communities: the role of earthworms, plant
functional group identity and seed size. Oikos
117:1026–1036
Estrada JA, Flory SL (2015) Cogongrass (Imperata cylindrica)
invasions in the US: mechanisms, impacts, and threats to
biodiversity. Global Ecol Cons 3:1–10
Flory SL, Clay K (2010) Non-native grass invasion suppresses
forest succession. Oecologia 164:1029–1038
Funk JL, McDaniel S (2010) Altering light availability to restore
invaded forest: the predictive role of plant traits. Restor
Ecol 18:865–872
Gaffney JF (1996) Ecophysiological and technological factors
influencing the management of cogongrass (Imperata
cylindrica). Ph.D. Dissertation, University of Florida,
Gainesville, FL
Gelman A, Hill J (2006) Data analysis using regression and
multilevel/hierarchical models. Cambridge University
Press, Cambridge
Gelman A, Carlin JB, Stern HS, Dunson DB, Vehtari A, Rubin
DB (2013) Bayesian data analysis. CRC Press, Boca Raton
Greene D, Johnson E (1994) Estimating the mean annual seed
production of trees. Ecology 75:642–647
Gurevitch J, Fox G, Wardle G, Taub D (2011) Emergent insights
from the synthesis of conceptual frameworks for biological
invasions. Ecol Lett 14:407–418
Hamilton MA, Murray BR, Cadotte MW, Hose GC, Baker AC,
Harris CJ, Licari D (2005) Life-history correlates of plant
invasiveness at regional and continental scales. Ecol Lett
8:1066–1074
Harper JL, Lovell P, Moore K (1970) The shapes and sizes of
seeds. Annu Rev Ecol Syst 1:327–356
Inderjit, Cadotte M, Colautti R (2005) The ecology of biological
invasions: past, present and future. In: De Koon H, Visser
EJW (eds) Invasive plants: ecological and agricultural
aspects. Birkhauser-Verlag AG, Basal, pp 19–44
Kolar CS, Lodge DM (2001) Progress in invasion
biology: predicting invaders. Trends Ecol Evol 16:199–
204
Lau JA, Schultheis EH (2015) When two invasion hypotheses
are better than one. New Phytol 205:958–960
Leishman M, Westoby M (1994) The role of large seed size in
shaded conditions: experimental evidence. Funct Ecol
8:205–214
Lippincott CL (2000) Effects of Imperata cylindrica (L.) Beauv.
(Cogongrass) invasion on fire regime in Florida sandhill
(USA). Nat Area J 20:140–149
Lloret F, Me
´dail F, Brundu G, Camarda I, Moragues E, Rita J,
Lambdon P, Hulme PE (2005) Species attributes and
invasion success by alien plants on Mediterranean islands.
J Ecol 93:512–520
Propagule quality mediates invasive plant establishment 2331
123
Lockwood JL, Cassey P, Blackburn T (2005) The role of
propagule pressure in explaining species invasions. Trends
Ecol Evol 20:223–228
Lockwood JL, Cassey P, Blackburn TM (2009) The more you
introduce the more you get: the role of colonization pres-
sure and propagule pressure in invasion ecology. Divers
Distrib 15:904–910
Loh RK, Daehler CC (2007) Influence of invasive tree kill rates
on native and invasive plant establishment in a Hawaiian
forest. Restor Ecol 15:199–211
Lonnberg K, Eriksson O (2013) Rules of the seed size game:
contests between large-seeded and small-seeded species.
Oikos 122:1080–1084
MacDonald GE (2004) Cogongrass (Imperata cylindrica)—biology,
ecology, and management. Crit Rev Plant Sci 23:367–380
Moles AT, Westoby M (2004a) Seedling survival and seed size:
a synthesis of the literature. J Ecol 92:372–383
Moles AT, Westoby M (2004b) What do seedlings die from and
what are the implications for evolution of seed size? Oikos
106:193–199
Moles AT, Hodson DW, Webb CJ (2000) Seed size and shape
and persistence in the soil in the New Zealand flora. Oikos
89:541–545
Patterson DT (1980) Shading effects on growth and partitioning
of plant biomass in cogongrass (Imperata cylindrica) from
shaded and exposed habitats. Weed Sci 28:735–740
Primack RB (1978) Regulation of seed yield in Plantago. J Ecol
66:835–847
Quinn LD, Holt JS (2008) Ecological correlates of invasion by
Arundo donax in three southern California riparian habi-
tats. Biol Invasions 10:591–601
Quinn LD, Holt JS (2009) Restoration for resistance to invasion
by giant reed (Arundo donax). Invasive Plant Sci Manag
2:279–291
Ramsey CL, Jose S, Miller DL, Cox J, Portier KM, Shilling DG,
Merritt S (2003) Cogongrass [Imperata cylindrica (L.)
Beauv.] response to herbicides and disking on a cutover site
and in a mid-rotation pine plantation in southern USA. For
Ecol Manag 179:195–207
Rejma
´nek M, Richardson DM (1996) What attributes make
some plant species more invasive? Ecology 77:1655–1661
Richardson DM, Rejmanek M (2004) Conifers as invasive
aliens: a global survey and predictive framework. Divers
Distrib 10:321–331
Ruiz GM, Fofonoff PW, Carlton JT, Wonham MJ, Hines AH
(2000) Invasion of coastal marine communities in North
America: apparent patterns, processes, and biases. Ann Rev
Ecol Syst 31:481–531
Schramm JW, Ehrenfeld JG (2010) Leaf litter and understory
canopy shade limit the establishment, growth and repro-
duction of Microstegium vimineum. Biol Invasions
12:3195–3204
Simberloff D (2009) The role of propagule pressure in biolog-
ical invasions. Annu Rev Ecol Evol Syst 40:81–102
USDA, NRCS (2005) The PLANTS Database. Data compiled
from various sources by Mark W. Skinner. National Plant
Data Center. http://plants.usda.gov. Accessed 22 Oct 2005
Valladares F (2003) Light heterogeneity and plants: from eco-
physiology to species coexistence and biodiversity. In:
Esser K, Luttge U, Beyschlag W, Hellwig F (eds) Progress
in botany. Springer, Berlin pp 439–471
Van Wilgen B, Siegfried W (1986) Seed dispersal properties of
three pine species as a determinant of invasive potential.
S Afr J Bot 52:546–548
Willard TR, Hall DW, Shilling DG, Lewis JA, Currey WL
(1990) Cogongrass (Imperata cylindrica) distribution on
Florida highway rights-of-way. Weed Technol 4:658–660
Williamson M (1996) Biological invasions. Chapman Hall,
London
Wills G (1975) Effect of light and temperature on growth of
purple nutsedge. Weed Sci 23:93–96
Wilson CH, Caughlin TT, Civitello DJ, Flory SL (2015) Com-
bining mesocosm and field experiments to predict invasive
plant performance: a hierarchical Bayesian approach.
Ecology 96:1084–1092
2332 J. A. Estrada et al.
123
A preview of this full-text is provided by Springer Nature.
Content available from Biological Invasions
This content is subject to copyright. Terms and conditions apply.