Content uploaded by Juliana Barrett
Author content
All content in this area was uploaded by Juliana Barrett on Mar 17, 2016
Content may be subject to copyright.
Seedling Recruitment Limitation in White Clover (Trifolium repens; Leguminosae)
Author(s): Juliana Panos Barrett and John A. Silander, Jr.
Source:
American Journal of Botany,
Vol. 79, No. 6 (Jun., 1992), pp. 643-649
Published by: Botanical Society of America
Stable URL: http://www.jstor.org/stable/2444880
Accessed: 14/12/2010 03:29
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=botsam.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Botanical Society of America is collaborating with JSTOR to digitize, preserve and extend access to American
Journal of Botany.
http://www.jstor.org
American Journal of Botany 79(6): 643-649. 1992.
SEEDLING RECRUITMENT LIMITATION IN WHITE
CLOVER (TRIFOLIUM REPENS; LEGUMINOSAE)1
JULIANA
PANOS BARRETT2 AND JOHN A. SILANDER, JR.
Department of Ecology
and Evolutionary Biology, The University of Connecticut,
Storrs,
Connecticut
06268
White clover
(Trifolium repens L.)
is
a stoloniferous clonal legume. Dynamics
of white clover
populations
are controlled
by the recruitment
of
both seedlings (genets)
and node production (ramets).
To understand the
role of
seedling dynamics
in natural populations of white clover,
the recruitment and mortality of seedlings in a lawn and
pasture clover population
were followed and the roles ofbiotic and
abiotic factors in seedling recruitment and survival were assessed.
Seedling recruitment
was several orders of magnitude less than
vegetative recruitment in both lawn and pasture populations.
However, at least
several new genets are added to these populations
on a yearly basis.
In
both locations, physical
disturbances may open
sites
for seed
germination,
as disturbed sites had
higher
seed
germination
rates than undisturbed
sites.
However,
these sites
were
not favorable for long-term establishment
due to such factors as frost upheaval. Therefore, there
are
trade-offs
in
safe
site
attributes
for
seed germination
and establishment that strongly limit recruitment
in
natural populations
of white clover.
Seed germination and survival
in
closed swards dom-
inated by vegetatively spreading
perennials are typically
very low (Sarukhan and Harper,
1973; Ennos, 198
1;
Lov-
ett Doust, 1981; Schmid,
1984;
Hartnett
and Bazzaz,
1985). Consequently, the
dynamics of these populations
are likely to
be determined
primarily by
the
growth
and
dispersal of vegetative
modules, and
sexual seed
produc-
tion
is
important only
in
the
colonization
of new
or
dis-
turbed sites.
If
this is
the
case,
then
one
can ask
what,
if
any,
role does seed
production
play
in the
dynamics
and
structure of
closed, clonally
dominated plant populations?
Moreover,
if
seedling
recruitment is rare, what conse-
quences does this have
for genetic variation
in
popula-
tions?
While
only
the first
question
is
addressed
in this
study, such questions
are
particularly germane to
white
clover
populations
where
genotypic
diversity
can be
high
(Burdon, 1980). Cahn and
Harper (1976) found an average
of
four leaf mark
phenotypes
per
10 cm2
sample.
Glidden
and
Saleem
(1985)
found
that approximately
90% of the
white clover ramets randomly
sampled from a natural
population
were
unique genotypes
up to an area of 10
M2.
It is therefore
of considerable
interest
to
assess
the role
that
seedling
recruitment
plays
in
the
structure and
dy-
namics of clonal plant populations.
Seed germination and establishment
are dependent on
safe
site
availability sensu
Harper (1977).
A
safe site offers
the
appropriate
biotic
and
abiotic conditions necessary
for seed
germination
and establishment.
Laboratory
and
greenhouse experiments (Harper, Williams, and Sagar,
' Received for
publication
8
August 1991;
revision
accepted
18 Feb-
ruary 1992.
The authors thank G. Anderson,
N. Barrett, M. Cain, R. Cook,
A.
W. H. Damman, T. Murray, and
R. Panos for valuable comments and
suggestions
on
this
manuscript.
This research was
supported
in
part by
NSF
grants
BSR-8616745 and BSR-8705747
to
JAS
and
S.
W.
Pacala
and
by
a Sigma
Xi
Grant-in-Aid
of Research. Facilities were
provided
by
The University
of
Connecticut
Agronomy Farm, Department
of
Ecology greenhouses,
and
Miss Isabelle
Atwood.
This
paper
was sub-
mitted
in
partial
fulfillment
of
the requirements
for
the Ph.D. degree
at
The
University
of Connecticut.
2 Author
for
correspondence,
current address:
The
Nature
Conser-
vancy, 55 High St., Middletown,
CT 06457.
1965; Ross and Harper,
1972; Hartgerink and Bazzaz,
1984) have
shown
that microsite availability, emergence
time, seedling density,
and environmental heterogeneity
can all affect seed germination
and seedling growth.
How-
ever, these studies account
for
neither natural
seed pool
recruitment limitations
nor the combined effects
of nat-
ural
field conditions
(temperature
fluctuation, predation,
competition,
overwintering ability)
on
germination
and
growth. These combined
factors may strongly affect
actual
seedling densities
in
the field. For
example,
seed
germi-
nation and early establishment
of white clover were
found
to
be
many
times
higher
in
an artificial lawn
in
the
green-
house (H. Reynolds,
University of Connecticut,
personal
communication)
than
in
a natural lawn population
(Pa-
nos, 1989).
Seedling recruitment
depends on seed number and
availability
of recruitment sites. For
white
clover,
most
findings to date have
been incidental to other
research
objectives, and
the
critical
field
experiments
have not been
done to examine
explicitly
the
dynamics
of
seedlings.
Turkington et al. (1979)
found that white clover
seedlings
failed to recruit
in
a
permanent
pasture except
on
specific
types
of disturbed sites such as
molehills.
Chapman
(1987)
also found
very
low levels of
seedling
establishment
for
white clover
in
grazed
sheep pastures, where
establish-
ment sites
were
not
associated
with
disturbances.
In
these
pastures,
the low
seedling
recruitment
of white
clover was
attributable to low
germination
rates
and
high
seedling
mortality
rather
than
lack
of
seeds
(Chapman
and
An-
derson, 1987).
Seedling
recruitment
(5.5 seedlings
per
m2
per year) represented
only
4.2% of the seeds dispersed
each summer. Far fewer
seedlings actually
became estab-
lished plants (4.4% of
seedlings). Ennos (1981)
similarly
found
very
low
or
negligible
levels of natural
seedling
recruitment.
The
purpose
of this study was to evaluate factors
lim-
iting
recruitment of white clover
in
field populations
by
1) following
the
dynamics
of natural seedling cohorts
over
a 1-yr period
to determine
recruitment
rates
and
2) ex-
perimentally
examining
the effects of
several
biotic site
attributes
(heterospecific
shading
and root mat
formation)
and an abiotic site
attribute (disturbance size)
on
seedling
recruitment.
643
644 AMERICAN JOURNAL OF BOTANY [Vol. 79
TABLE 1. Treatments used to
examine the
effects
of biotic
and abiotic
factors on
seedling recruitment
N = 6a
Artificially
seeded
(35 seeds/ Naturally
plot) seeded
Trt
A
aboveground
vegetation
removed
25-cm-diam L,
G L, G
Trt B
above- and
belowground
vegetation
removed
25-cm-diam L, G, U L, G, U
Trt C above- and
belowground
vegetation
removed
l0-cm-diam L, G, L, G
Trt
D natural
vegetation-undisturbed
25-cm-diam L,
G L,
G
a L = lawn; G = grazed pasture;
U = ungrazed
pasture.
MATERIALS AND METHODS
The
species-White
clover
(Trifolium
repens L.) is a
widespread,
stoloniferous
perennial
found
in
grasslands
and
open
pasture
throughout
the
temperate
zone.
This
Eurasian
species is fully
naturalized
in eastern North
America
(Fernald, 1970).
The
natural
history,
dynamics,
and
growth
patterns
of
white clover have been well
doc-
umented
(Erith,
1924;
Burdon,
1983;
Turkington
and
Burdon, 1983).
A broad
range of
temperature
and light
regimes
produces/stimulates germination
(Turkington and
Burdon, 1983).
Seeds
generally
germinate
in
the
soil with-
in 24 hr
at 18-20 C
(Erith,
1924).
Seeds are
easily shaken
from
the pod
(Erith,
1924) and are
passively
dispersed
locally. Seeds
may
also be
locally
distributed by livestock
(Burdon, 1980) and
other
animals.
Upon
germination,
white
clover
seedlings
develop
a
primary, erect stem
with
a series of
primary,
trifoliate
leaves. After
seven to
eight
primary
leaves have
developed
fully,
lateral buds
in
the
axils of
the
older
leaves
may begin
to
produce
primary
or
lateral stolons.
These stolons are
composed
of
modules.
(A
module is the
iterated unit
of
growth
forming a colony
or clone
(Jackson,
Buss, and
Cook,
1985)).
For white
clover,
the
module
consists
of
a
node
and
the
associated
internode,
with the
apical
mer-
istem
iterating
new
modules. Once
a
node comes
in
con-
tact
with
the soil
surface,
roots
may
form.
Internode se-
nescence,
and thus module
independence, occurs
over
a
temporal
scale of from
1
to 6 mo of
root initiation
(H.
Reynolds and
J.
A. Silander,
University
of
Connecticut,
unpublished
data)
and is also
genotype dependent
(J.
P.
Barrett
and J. A. Silander,
University
of
Connecticut,
unpublished
data).
Study
sites-Representatives
of the
two main com-
munity
types
that
may
be dominated
by
white
clover
were
selected
for
study:
a mown
lawn and a grazed
pasture
in
northeastern Connecticut. These
sites
contrasted
by
show-
ing
differences
in
disturbances,
white
clover
density,
and
species
composition.
The soil of both
locations is a typic
fragiochrept (Ilgen
et
al.,
1966).
The
lawn has
existed as
such for
at least 20
yr.
It was mown at
2-
to 3-wk intervals
during
the
study.
Natural disturbances included worm
casts and ant
hills,
but
with little mole
activity.
The
pas-
ture has been used for
agricultural
purposes
for
at least
90
yr.
Dairy
cattle
grazed
the
pasture
daily.
Natural dis-
turbances
included
hoof
prints
and
trampling,
cow
patty
deposition,
worm
casts,
ant
hills,
and
mole
activity.
Natural
seedling
and
module
recruitment-To
quantify
seedling recruitment
and
vegetative node
production,
ten
randomly
located,
10
x 10-cm2
plots were
established
in
both the
lawn
and
pasture.
In these
plots,
white
clover
stolons were
censused
weekly
for
node
production. In
addition, ten
neighboring quadrats
were
censused
for
seedling recruitment
for a total of 20
plots
per
site.
This
census was
conducted from
May
through
July 1986.
Seedling dynamics
Based
on
the
1986 results of
seed-
ling
recruitment, a larger
plot size
was
determined nec-
essary to
characterize
seedling
recruitment.
Therefore a
single
permanent
plot
was
set
up
in
both the
lawn and
pasture
and
censused
weekly
for
seedlings.
The
rectan-
gular areas
surveyed
in
the
lawn,3
m2, and
in
the
pasture,
42.24
i2, were
based on
pilot
studies
of
1986
and
reflect
differences
in
seedling
occurrence
for the
two
locations.
Seedlings were
censused
weekly
from
May
through
Oc-
tober 1987
(the
full
growing
season for
clover
in
Con-
necticut),
and
all
seedling
cohorts
that
recruited
in
that
time
interval
were
marked
and followed at
each
census.
Both
study
areas
were
again
censused
in
April
and
May
1988
to
determine
overwintering
success
of
seedlings
re-
cruited
in 1987.
Competitive interactions-To evaluate the effect
on
seedling recruitment
of shoot
and root
interference,
we
established
artificially and
naturally
seeded
circular
plots
in
the lawn
and
grazed
pasture
(Table 1).
In
Treatment
(Trt)
A,
aboveground
vegetation
other
than
white
clover
seedlings was clipped
on a weekly basis to soil level
throughout
the
growing season
to
evaluate
light
interfer-
ence from
shading.
In Trts
B and
C both above-
and
belowground
vegetation were
removed.
Roots were ex-
cavated,
and soil
was removed
from the
roots and
replaced
in
the
plots.
This
mimicked
disturbance
from
frost
up-
heaval
and
certain
animal
disturbances
(i.e.,
mole
hills,
ant
mounds,
etc.).
Vegetation
other
than white
clover
seedlings
was
clipped
on
a weekly
basis
throughout
the
course of the
experiment.
To remove the
possible
effects
of
grazing
and
trampling
on
seedling
recruitment,
Trt
B
(both
artificially
and
naturally
seeded)
was set
up in
a
section of
pasture
protected from
cows, as
well
as
in
the
grazed
pasture.
The
control
plots
(Trt
D) were
left
un-
disturbed. Each
treatment was
replicated
six
times.
Disturbance
scale-Two plot sizes reflecting
distur-
bance levels
similar
to those
caused
by
moles and
cows
were
superimposed on Trts
B
and
C (Table
1).
Pilot
studies
performed
in
1986
indicated
that
very small
disturbances
(on
the
order
of 2 cm
in
diam)
showed
no effect on
module
or
seedling dynamics
(J.
P. Barrett,
unpublished
data).
However,
larger
disturbances such
as cow
hoof
prints
range
in
size
from 10
to 15 cm
and
mole
disturbances
(up
to
25-cm-diam)
do
potentially
affect
both
module and
seedling
dynamics.
Trts
B and
C, with
the
same
biotic
removal
treatment,
were
compared
on
the
basis of size
of
disturbance-
25 cm
and
10
cm
diameter
disturbances,
respectively,
such
that
the
effects
of
disturbance
scale
on
seedling
dynamics
could be
examined.
Trts
A
and
D were
maintained as 25-cm-diam
disturbances.
In
artificially
seeded
plots,
35
white
clover
seeds were
June 1992] BARRETT AND SILANDER-CLOVER SEEDLING RECRUITMENT 645
6
5_
E
(D 4
(/1
C 3_t\
e 2
(0
1 5 6 7 8 9 10 11 12 13 15 16 17 18 20 47 49
Time (Weeks)
Fig.
1. Densities
of
naturally
occurring
seedlings
in
the
pasture
(open
circle)
and
lawn
(closed
circle)
during
the
period
8 May
1987-2
May
1988.
randomly
scattered
over
each
treatment
plot.
This
value
represents
an
intermediate
level
of
the
natural
seedling
pool
found
in the
lawn
and pasture
(lawn-eight
viable
seeds per
m2
per
5 cm
depth;
pasture-62
viable
seeds
per
m2
per 5 cm
depth;
Panos,
1989).
In addition,
35
seeds
per
plot
(whether
10- or
25-cm-diam)
is
below
the
level
at
which
density-dependent
factors
affect
seedling
survival
(H. Reynolds
and
J.
A. Silander,
unpublished
data).
All
plots
(treatments
and
controls)
were
lightly
covered
with soil
to maximize
germination.
This
is a
natural
phe-
nomenon
that
occurs
when
worm
casts,
ant
hills,
and
freeze-thaw
cycles
cover
seeds
or
incorporate
them
into
the soil.
New
seedlings
were marked
and
their
fates
fol-
lowed
weekly
from June
through
October
1987
and
again
in April
and
May 1988.
The total
number of
plots
was
two locations
x four treatments
x two seeding
regimes
x six replicates
plus
the subset
of
plots
in the
ungrazed
pasture
(one
treatment
x two seeding regimes
x six
rep-
licates),
or 108
plots.
RESULTS
Natural
seedling
and
module
recruitment-The
mean
numbers
of new
seedlings
per
m2 in
the
lawn
(3.3) and
pasture
(3.5)
were
three to four orders
of
magnitude
less
than
the
mean
number
of
modules
recruited (or
ramets
as
intemodes
senesce)
per
m2-lawn:
10,750;
pasture:
14,800
for the
1986 census
from
May
through
July.
Given
that
seedling
recruitment
was
very
low, but
did occur,
it is
important
to determine
the
likelihood
of seedlings
re-
cruiting
to the adult
class, and
thereby
the extent to
which
seedlings
contribute
to
the
dynamics
and
the
genotypic
structure
of
the
population.
In
the
pasture,
total
seedling
density
was
greatest
in
the late
spring
of
both
1987
and
1988 (2.4
and
2.5 per
m2,
respectively),
and steadily
de-
creased
throughout
the
summer
of
1987
(Fig.
1).
Only
0.2 seedlings
per
m2
overwintered
in the
pasture;
this
represented
67% of
the seedlings
still
alive
in
the
fall.
The
lawn
seedlings
showed a very different
pattern.
Here
the
greatest
number
of seedlings
was
in
the early fall
(5.7
per
m2), twice that
found
in
the springs
of
either 1987
or
1988
(2.3
and
1.7 per
M2, respectively).
The majority
of seed-
lings (67%) found
in September
and October
were on
worm cast
mounds,
while none
were found
on
worm cast
mounds
in
the
pasture.
The number
of lawn
seedlings
105
r J(14)
4A
(103)
(48)
1007 B(55) 5
W 25 o
0
C\
-o 20 --
cJ)
15 _
a
E A
10 0
Z IQ~ 00?-00oN? O -O O
0 L-' ' ' ' , , - - H! - 5-+
518 5115 6110 6O17 711 7110 7/15 7122 7/31 815 8117 913 9/11 9/22 1011
10/13 5/2 519
Time (mo./d.)
Fig. 2. Number
of naturally
occurring
seedlings
in pasture
cohorts
from
May
1987 through
May
1988.
Each letter
represents
a cohort
of seedlings.
The
same
plot was
searched
for seedlings
each
week.
646 AMERICAN JOURNAL OF BOTANY [Vol. 79
D
cn~~~~~~~~~~~~~~
V)~~~~~~~~~~~~~~~
08
._ A\
a) cE_
?' 6 0\ 0
E)
a 2 -A A A A Az A
Z B
0 , I ., I . I
, , ., . I I
6/11 6/18 7/1 7/9 7/16 7/22 7/30 8/5 8/17 9/3 9/11 9/21 10/2 10/9 4/25 5/9
Time (mo./d.)
Fig. 3. Number of naturally
occurring seedlings in lawn cohorts
from June 1987 through May 1988. Each letter represents
a cohort of
seedlings.
The same plot was searched
for seedlings each week.
that survived
the
winter was 2.3 per
m2, representing
58%
of the
seedlings
still alive
in
the fall.
Seedling
dynamics-
Figures
2 and
3 follow the
fate of
seedling
cohorts
that successfully recruited
on the plots
over
1
yr.
New cohorts
in the pasture
appeared six
times
from May
through August
(Fig. 2). Seedlings
were
found
on
yr-old
cow patties,
hoof prints as
well as sites
with no
apparent
disturbances.
In
May,
the
pasture
showed
a
large
burst
of
seed germination
(15.8
seedlings m-2).
However,
mortality
was very high
and rapid- 50%
died within
the
first week
of growth.
The second cohort
(8.5 seedlings
m-2) also
had high mortality
the first
month (84%).
Sub-
sequent
cohorts comprised
fewer than three seedlings
m-2
until the following spring.
Significantly,
those seedlings
that
survived the
winter
were the
earliest cohorts that
had
become well established
by
fall. Most of
the
surviving
seedlings
(approx. 67%)
were
growing
next
to or on
yr-
old
cow
patties.
In
the
lawn (Fig. 3),
only two cohorts
occurred
from
May
through August.
In
September
and October,
the ap-
pearance
of
three
new cohorts coincided
with
an
increase
in
the prevalence
of
worm
cast mounds
(67%
of
seedlings
were on
mounds).
Several members
ofthe
first two cohorts
(A and B) survived
the winter (14% and 100%
of the
cohort
size, respectively),
while for
later
cohorts,
only
the
two that
appeared
in
early
to mid-September
had
any
survivors
(17% and 33% of
the
initial cohort
size,
re-
spectively).
The
last cohort to appear
in
the fall failed
to
survive
the winter.
Competitive
interactions and disturbance
scale-This
experiment
was set
up
to determine the effects
of biotic
site attributes (shading
and
root
mat
formation) and the
size
of disturbance
on
seedling growth
and
survival. Nat-
urally
seeded
plots
in the grazed pasture showed very
low
seedling
recruitment and were
significantly
different from
supplemental seeding
plots: using
Mann-Whitney Test:
Trt A: W = 45.0,
P < 0.01; Trt B: W = 57.0,
P < 0.005;
Trt C: W
= 45.0, P < 0.01; Trt D: W
= 53.5, P < 0.05.
A
total
of
one seedling
was
found
in
each of the naturally
seeded
plots
for treatments A, B,
and C. Trt
D (undis-
turbed natural
vegetation)
had a total
of five
natural re-
cruits.
All
but
one of these
eight seedlings
died
by
the end
of
September.
The
single exception
failed to survive
the
winter. This outcome
suggests strong seedling
recruitment
limitations
for white clover
in
open,
disturbed sites
in
pastures.
Results
of the
grazed
pasture disturbance
experiment
with supplemental
seedings are shown
in
Fig. 4. Undis-
turbed
vegetation
showed the lowest
seedling
recruitment
of
the
four treatments. Significant differences
in
seedling
recruitment were
found between disturbance
size treat-
ments during the
field season (W
= 21.5,
P < 0.01). More
seedlings
recruited to and survived
in
the
1
0-cm-diam
plots than
in
the
25-cm-diam plots, indicating
significant
differences
in
the
scale
of
safe sites created by disturbance.
However,
the
type
of disturbance
(removal
of above-
ground vegetation
vs. removal of both above- and be-
lowground vegetation)
showed
no
significant
differences
in
recruitment and
survival throughout
the growing sea-
son
(P > 0.05).
No seedlings
survived the
winter
in
any
of these
plots.
This
again
reflects
strong
seedling
recruit-
ment limitation.
Recruitment
of
seedlings
in
the
naturally
seeded
plots
was much higher
in
the lawn (19 total)
than
in
the pasture.
June 1992] BARRETT AND SILANDER-CLOVER SEEDLING RECRUITMENT 647
35 -
:' 30
25 -
Q)T
0 20
* 15
11-4
0 -
-
* T
0 2 4 6 8 10 12 14 16 44 45
Time (Weeks)
Fig. 4. Changes in
the mean
number of
seedlings per
seeded treat-
ment
in
the pasture
? SE: 10-cm-diam, above-
and
belowground veg-
etation
removed
(square);
25-cm-diam,
aboveground vegetation re-
moved
(open circle);
25-cm-diam,
above- and
belowground
vegetation
removed
(closed
circle); natural
vegetation
(triangle) during
the period
9
June 1987-2
May
1988.
N= 6.
However,
as was the case
in
the
pasture,
the number of
natural
recruits
was
significantly lower
than
the recruit-
ment
in
supplementary
seeded
lawn
plots (Trt
A: W =
51.0, P < 0.01; Trt B: W = 57.0, P < 0.005; Trt C: W
= 57.0, P < 0.005; Trt D: W = 44.5, P < 0.01). Of these
19
seedlings: five
were
found
in
Trt
A
(aboveground veg-
etation
removed-25-cm-diam),
five were
found
in
Trt
B
(above- and
belowground
vegetation
removed-
25-cm-
diam),
two
were found
in
Trt C
(above-
and
belowground
vegetation
removed-
10-cm-diam), and
seven
were found
in
Trt
D (undisturbed lawn).
Only one
seedling
in
Trt A
survived the
winter to become
an
established
genet.
Again,
this result
indicates recruitment
limitations
in
lawn
pop-
ulations
with
natural seed rain.
Figure
5
shows a pattern
of
seedling
recruitment
(from
supplemental
seeding)
and
mortality
in
the
lawn
similar
to that found
in
the
grazed pasture.
The notable
exception
is that
seedling
recruitment
in
the lawn was
higher than
35
" 30
0)
u) 25
I_ T T
20 i
15TT
0 '
0 2 4 6 8 10 12 14 16 44 45
Time (Weeks)
Fig.
5. Changes
in
the mean
number
of
seedlings per
seeded treat-
ment
in
the lawn
? SE:
1
O-cm-diam,
above- and
belowground
vegetation
removed
(square);
25-cm-diam,
aboveground
vegetation removed
(open
circle);
25-cm-diam,
above- and
belowground vegetation
removed
(closed
circle);
natural
vegetation
(triangle) during
the
period
9 June
1987 to
25 April 1988. N = 6.
35 - 0
~"30 IlK.
25-
'' 20 - l
20
?- I'1- <r
15 h
5 1< [ |<3_; ;
0
0 2 4 6 8 lo 12 14 16 44
45
Time ( Weeks )
Fig. 6. Changes
in
the
mean
number
of
seedlings per
seeded treat-
ment
? SE (25-cm-diam,
above-
and
belowground
vegetation
removed)
in
the:
ungrazed pasture
(open
circle), grazed
pasture
(closed
circle),
lawn
(square) during
the
period
9 June 1987-25 April
1988. N = 6.
in the
pasture
for all treatments. The undisturbed
lawn
vegetation
showed
the
lowest
level of
recruitment,
while
recruitment
and survival
were
greatest
for
Trt
C (I10-cm-
diam plot)
with
significant
differences
occurring
between
this
treatment
and the
control
(W = 5 1.
0,
P < 0.
01). As
in the
pasture,
no significant
differences
were
found
be-
tween
Trts A and
B,
indicating
that
the
presence
of a root
mat has little or no effiect
on seedling
recruitment
and
survival.
All
treatments
except
the
undisturbed
lawn
veg-
etation
(Trt
D) and to a lesser
degree
Trt
A, with
only
aboveground
vegetation
removed,
had high
winter
mor-
tality
(Trt
A-25% mortality;
Trt
B-56%; Trt
C-44%;
while
Trt D was only
18%). However,
seedling
numbers
by
the
end of
the
season had become too low overall
to
show
significant
differences in
winter
mortality.
In
both
the lawn
and
pasture,
removal
of
aboveground
vegetation
significantly
increased
seedling
recruitment
in
comparison
with
the
undisturbed
vegetation
(lawn:
W =
5
1.0,
P < 0.0
1; pasture:
W = 42.5, P < 0.05). However,
the
presence
of
an undisturbed
canopy
and/or
root
mat
benefits
the
overwintering
success of
seedlings,
particu-
larly
in
the lawn
(Fig.
5). Overall,
the
scale of
disturbance
did have a major effect
on seedling
recruitment
and es-
tablishment in both locations,
with
the smaller
distur-
bances showing greater
recruitment
and survival
during
the
growing
season than
the
larger
disturbances.
While
no seedlings
survived
the winter
in the pasture,
more
seedlings
survived
the
growing
season in
the
smaller
dis-
turbances
(Trt
C- 56%)
than
in
the
equivalent
larger
ones
(Trt
B-44%). However,
this
difference was not
signifi-
cant.
The presence
of
surrounding
vegetation
and
distur-
bance scale together present
a trade-off in seed germi-
nation
and
establishment.
While an
intact
canopy
decreases
the
probability
of
seedling
recruitment,
such
conditions
increase
the
probability
of a seedling
surviving
the
winter.
To determine the
effiect
of
protection
from
grazing
and
trampling
on
seed
germination
and'survival,
Trt
B
(above-
and
belowground
vegetation
removed
-2 5
-cm-dianm)
was
replicated
in an ungrazed
section
of the
pasture.
These
results
along
with
those
from
the
parallel
treatment in
the
grazed
pasture
and lawn
are
shown in
Fig.
6. Mean seed-
ling gzermination
was,
98% in the-
ungrazed npasitur
com-
648 AMERICAN JOURNAL OF BOTANY [Vol. 79
pared with 57%
in
the
grazed
pasture
and 80%
in the lawn.
The number of established
seedlings
in
the ungrazed
pas-
ture
site was
significantly
greater
than in the
grazed
site
during the early part
of the
growing
season
(W
= 21.0,
P
< 0.01). However,
no seedlings
survived the
winter
in
the
ungrazed pasture
site, negating any
net
differences
between sites attributable
to cattle.
DISCUSSION
The recruitment of
seedlings reported here
on both
disturbed
and undisturbed
sites,
in
both the
lawn and
pasture,
contrasts sharply
with other studies
in which
seedling
recruitment was
confined
to
disturbed
sites
(Cahn
and
Harper, 1976;
Turkington
et
al., 1979; Ennos,
1981).
While seedling recruitment
of white clover may
increase
on disturbed
sites,
this does not
ensure establishment.
The large increase
in
seedling recruitment
in
the
lawn
in
early
fall
may
be attributed
to
earth
worm
activity.
The
production of
wormcasts
in
the temperate zone
reaches
a maximum
in
early spring
and
again
in
the autumn and
is typically
low or
absent
at other times
of
the
year
(Evans
and
McL. Guild,
1947).
In spite
of
the
importance
of
worm cast mounds as germination sites,
it
is surprising
that
none
of such
seedlings
in
this
study
survived to adults.
Rather, those seedlings
that did
survive
1
yr
in
the
lawn
could not be
identifiably
associated
with
any
natural dis-
turbance.
In the pasture, cows
are responsible for creating
dis-
turbances
(e.g.,
cow
patties
and hoof
prints)
in which
germination
is higher,
but
they
are also
responsible
for
high seedling mortality
through grazing and trampling
(J.
P. Barrett, personal observation).
Cow patties are
a po-
tential safe site
for
both
seedling germination
and estab-
lishment as cows
graze
within only 20 cm
of
such
sites
over a period
of at least
1
yr (Norman
and
Green,
1958;
Castle and
MacDaid, 1972). We found patty sites
to be
the most
successful sites
for
recruitment
of
seedlings
to
the adult
stage.
Differences
in
disturbance
size and the effects of shading
do influence recruitment success.
Seedling
recruitment
and
survivorship
were much
greater
for the 10-cm dis-
turbance than for the
25-cm
disturbance
in
both the
lawn
and
pasture (Figs. 4, 5).
This
finding, together
with
pilot
study findings
on
module
dynamics (in
which
interme-
diate level disturbances
[7-cm-diam gaps]
led
to
greater
lateral branch
production
than small scale disturbances
[2-cm-diam gaps; J. P. Barrett, unpublished data]),
in-
dicates
that intermediate level
disturbances
may have
more favorable effects on white clover
growth
and seedling
recruitment.
For
seeds,
relatively small disturbances
(10-
cm-diam)
elicited
higher
germination rates
than larger
ones
(25-cm-diam), probably
due
to the more favorable
temperature
and
humidity
microclimate.
In particular,
beneficial effects
of
shading
are
apparent
in
the
ungrazed
pasture treatment (Fig.
6). The highest germination
rate
occurred
in
the
ungrazed
pasture (tallest vegetation),
fol-
lowed by the lawn and the
grazed pasture (shortest
stature
vegetation). However,
the effects of shading may
become
detrimental once
the seedlings start to grow and
light (or
other
resources)
becomes limiting. Ross and Harper
(1972)
found
that the
growth
rate of a seedling was positively
correlated
with
the three-dimensional space available
to
it.
They suggest
that
the potential
for resource
capture
by
a seedling
is determined
by the
number
and location
of
individuals
already
using
the resources.
Similar
studies
show
that survival
in
Viola
seedlings
is
related
inversely
to neighboring
adult
density
(Waller,
1981).
Schellner,
Newell,
and
Solbrig
(1982)
also showed
that the
negative
effect
of
adult
Viola
plants
on seedling
survival
is due to
competition
for common
resources
(such
as light).
Sim-
ilarly,
the scale
of
disturbance
in
relation
to both
water
stress
and
competition
with neighboring
individuals
ap-
pears
to be
an important
component
of
white
clover safe
sites.
Survival and establishment
of
white
clover
seedlings
over
the
winter also
depends
on
safe site
attributes
and
on
plant
size. The
only
seedlings to
survive
over the
winter
in
the
pasture
were
from
the early
spring cohorts
and were
usually
located on
or near
cow
patties.
The
majority
of
surviving
seedlings
in
the
lawn grew
in
the
turf
mat,
not
on worm
cast
mounds
which
are
subject to
frost
upheaval.
Markers placed
in worm cast mounds
to
locate
seedlings
were
completely
overturned
the
following
spring,
while
this
never occurred
in the turf
mat. Those plots
with
belowground
vegetation
removed
were also
more
prone
to frost
upheaval
as evidenced
by
marker and seedling
movement,
suggesting
that
the turf
mat
provides
protec-
tion from frost
upheaval
of the
soil
which would
expose
the roots and desiccate
seedlings.
Regehr
and Bazzaz (1979)
also
found
uprooting
of
plants
due
to frost
heaving
was
the
major
cause
of
seedling
mortality
for
Erigeron
can-
adensis.
Seedling
size as well as safe site
characteristics appear
critical
to survival.
In
both the pasture
and lawn, seeds
that germinated
in
the late
fall failed to become
sufficiently
well established
to
survive
the
winter.
Survival
of seed-
lings
as a function
of plant
size
has been
found
for many
species.
For
example,
Solbrig,
Newell,
and
Kincaid
(1980)
found that mortality
in
Viola sororia
seedlings
was
higher
among
the smallest
individuals
at
any given
time. They
suggest
that
individual
plant
size
in V.
sororia
is deter-
mined through
competition
for environmental
resources
such as
water, light,
and
space.
For white
clover,
the
same
is true;
competition
for resources
may affect
growth
and
establishment.
Seedling
size,
based
on both
the
effects
of
competition
and
time of
seedling
emergence,
thus affects
overwintering
success.
In
conclusion,
module recruitment
is
far
greater
than
seedling
establishment,
but
at least some new
genets
are
added at
a low rate to
both the lawn
and
pasture
popu-
lations on
a
yearly
basis.
Even
though
seed
bank
numbers
are
much lower
for the lawn than
for the
pasture,
differ-
ences
in
safe site
availability
result
in
a tenfold
greater
establishment
rate
of
seedlings
in
the
lawn than
in
the
pasture.
In
both
locations,
physical
disturbances
may open
sites
for
seedling
recruitment,
but these
sites
are
not
nec-
essarily
favorable
for
long-term
establishment.
Disturbed
sites
in
both
the
lawn
and
pasture
have
higher
seedling
recruitment
rates than
undisturbed
sites. However,
es-
tablishment rates
are
higher
in
undisturbed sites.
Thus,
there
is
a
trade-off
between site
attributes
for recruitment
and establishment.
Overall,
the
number
of successful
white
clover recruits
into closed
swards of
both the
lawn and
pasture
is
limited,
and
it
remains
to
be seen
whether
these
recruits
are sufficient
to balance
clone
mortality
over time.
June
1992] BARRETT AND SILANDER-CLOVER SEEDLING RECRUITMENT 649
LITERATURE CITED
BURDON, J. J. 1980. Intra-specific
diversity
in
a natural
population
of
Trifolium
repens. Journal of
Ecology 68: 717-736.
. 1983. Biological flora
of the British Isles. Trifolium repens
L.
Journal of Ecology
71: 307-330.
CAHN, M. G., AND J. L. HARPER. 1976. The biology
of
the
leaf
mark
polymorphism
in
Trifolium
repens
L. 1.
Distribution of
phenotypes
at a local scale. Heredity 37: 309-325.
CASTLE, M. E., AND E. MAcDAID. 1972. The decomposition
of
cattle
dung
and its effect on pasture.
Journal
of
the
British Grassland
Society
27: 133-137.
CHAPMAN,
D. F. 1987. Natural
re-seeding
and Trifolium repens
de-
mography
in grazed
hill
pastures.
II. Seedling
appearance and sur-
vival. Journal
of Applied Ecology
24: 1037-1043.
,
AND C. B. ANDERSON. 1987. Natural
re-seeding
and Trifolium
repens demography
in grazed
hill pastures.
I. Flowerhead appear-
ance and fate,
and seed
dynamics.
Journal of
Applied Ecology
24:
1025-1035.
ENNos,
R. A. 1981. Detection of selection in populations
of white
clover
(Trifolium
repens L.). Biological Journal
of
the Linnean So-
ciety 15:
75-82.
ERITH, A. G. 1924. White clover
(Trifolium repens
L.). A monograph.
Duckworth,
London.
EvANs, A. C.,
AND
W.
J.
McL. GUILD. 1947. Studies
on the relationship
between earthworms
and soil fertility.
I. Biological
studies in the
field.
Annals
of Applied
Biology
34: 307-330.
FERNALD, M. L. 1970. Gray's
manual of
botany.
D. Van Nostrand,
New York,
NY.
GLIDDEN, C., AND M. SALEEM. 1985. Gene flow
in Trifolium repens-
an expanding
genetic neighborhood.
In P. Jacquard,
G. Heim,
and
J.
Antonovics [eds.],
Genetic differentiation
and
dispersal
in
plants,
NATO, ASI Series
G: ecological
sciences,
vol. 5,294-309. Springer-
Verlag,
Berlin.
HARPER, J.
L. 1977. Population
biology
of
plants.
Academic Press,
London.
, J.
T. WILLIAMS, AND G. R. SAGAR. 1965. The behavior of
seeds in soil. I. The heterogeneity
of soil surfaces and its role
in
determining
the establishment of
plants
from seed.
Journal
ofEcol-
ogy
53: 273-286.
HARTGERINK,
A. P.,
AND F. A. BAZZAZ.
1984. Seedling-scale
environ-
mental
heterogeneity
influences individual fitness and population
structure.
Ecology
65: 198-206.
HARTNETT,
D. C., AND F. A. BAzzAz. 1985. The integration
of neigh-
borhood
effects
by
clonal
genets
in
Solidago canadensis. Journal
of
Ecology
73: 415-427.
ILGEN, L. W., A. W. BENTON, K. C. STEVENS, JR., A. E. SHEARIN, AND
D. E. HILL. 1966. Soil survey of Tolland County, Connecticut.
U.S. Department of
Agriculture, Soil Conservation Service, series
1961, No. 35. U.S. Government Printing Office, Washington, DC.
JACKSON, J. B. C., L. W.
Buss,
AND R. E.
COOK
[eds.]. 1985. Population
biology
of clonal organisms. Yale University Press, New Haven,
CT.
LovErr DousT, L. 1981. Population
dynamics and local specialization
in a
clonal perennial
(Ranunculus repens). I. The dynamics oframets
in
contrasting habitats.
Journal of Ecology 69: 743-755.
NoRMAN,
M. J. T., AND J. 0. GREEN. 1958. The local influence of
cattle
dung
and urine
upon
the
yield
and botanical
composition
of
permanent pasture. Journal
of
the British
Grassland Society 13: 39-
45.
PANOS, J.
M. 1989. Variation in
the architecture of white
clover (Tri-
folium repens L.) in contrasting
habitats and the
consequences
of
this variation on clonal
dynamics. Ph.D. thesis,
The University
of
Connecticut.
Storrs, CT.
REGEHR, D. L., AND F. A. BAzzAZ. 1979. The population dynamics
of
Erigeron canadensis,
a successional
winter annual. Journal of
Ecology 67: 923-933.
Ross, M. A., AND J. L. HARPER. 1972. Occupation of biological space
during seedling establishment. Journal
of Ecology
60: 77-88.
SARuKHAN, J., AND J.
L. HARPER. 1973. Studies
on
plant demography.
Ranunculus repens L.,
R. bulbosus
L. and
R. acris L. 1.
Population
flux
and survivorship. Journal
of Ecology
61: 675-716.
SCHELLNER, R. A., S. J.
NEWELL, AND 0. T. SOLBRIG. 1982. Studies
on the population biology of the
genus Viola.
IV. Spatial pattern
of
ramets and seedlings
in three stoloniferous
species. Journal of
Ecology 70: 273-290.
SCHMID, B. 1984. Life histories
in clonal plants
of the Carex
flava
group. Journal of Ecology
72: 93-114.
SOLBRIG,
O. T., S. J. NEWELL,
AND
D. T. KINCAID. 1980. The
population
biology
of
the
genus Viola. I. The demography
of Viola sororia.
Journal
of Ecology
68: 521-546.
TURKINGTON, R., AND J. J.
BURDON. 1983. The biology
of
Canadian
weeds. 57. Trifolium repens
L. Canadian Journal of Plant Science
63: 243-266.
- M. A. CAHN, A. VARDY, AND
J.
L. HARPER. 1979. The
growth,
distribution and neighbor
relationships
of Trifolium repens
in a
permanent pasture. III. The establishment and growth of Trifolium
repens
in
natural and perturbed sites. Journal
of Ecology
67: 231-
243.
WALLER, D. M. 1981. Neighborhood
competition
in several violet
populations. Oecologia 51: 116-122.
