Directory UMM :Data Elmu:jurnal:T:Tree Physiology:Vol16.1996:

Tree Physiology 16, 923--932
© 1996 Heron Publishing----Victoria, Canada

Hay-scented fern (Dennstaedtia punctilobula (Michx.) Moore)
interference with growth of northern red oak (Quercus rubra L.)
seedlings
JONATHAN LYON1,2 and WILLIAM E. SHARPE1
1

School of Forest Resources, Environmental Resources Research Institute, The Pennsylvania State University, University Park, PA 16802, USA

2

Present address: Department of Natural Science, Edgewood College, 855 Woodrow Street, Madison, WI 53711, USA

Received October 25, 1995

Summary We assessed the impacts of hay-scented fern
(Dennstaedtia punctilobula (Michx.) Moore) and subsoil liming (CaO amendments) on root and shoot growth of greenhouse-grown, first-year, northern red oak (Quercus rubra L.)
seedlings. Red oak seedlings and ferns were grown in reconstructed soil profiles of four common Pennsylvanian forest
soils. When grown in the presence of hay-scented ferns, with

or without subsoil liming, red oak seedlings had significantly
reduced height growth, and foliar, stem and total root biomass.
Fern foliar biomass was significantly reduced when ferns were
grown with red oaks, but there was no significant difference in
total belowground biomass of ferns. Belowground fern
biomass was concentrated in the upper soil profile, whereas red
oak roots showed a variety of distributions. In the presence of
ferns, fine root branching in red oak was reduced in the organic
horizons of three of the four soils tested. In both the presence
and absence of ferns, root branching in red oak was also
significantly and negatively correlated with the concentration
of 0.01 M SrCl2-extractable aluminum in the mineral horizons
(r 2 = 0.77). Subsoil liming generally improved root branching
in red oaks. The presence of ferns significantly reduced ectomycorrhizal infection frequency in red oak. We conclude that
hay-scented fern inhibited root branching and suppressed
above- and belowground biomass accumulation of first-year
northern red oak seedlings.
Keywords: root competition, soil acidity, subsoil liming.

Introduction

Hay-scented fern (Dennstaedtia punctilobula (Michx.)
Moore) is a rhizomatous, invasive, gap-colonizing species that
grows across wide gradients of edaphic and light conditions
(Conard 1908, Cody et al. 1977, Burnside 1988, Brach et al.
1993), and exhibits great phenotypic plasticity above- and
belowground (Horsley 1984, Hammen 1993). Dense ground
covers of hay-scented fern have been implicated in reducing
hardwood seedling growth and causing regeneration delays
and failures in Allegheny and northern hardwood forests
(Horsley and Marquis 1983, Maquire and Forman 1983, Drew

1988, 1990, Horsley 1993a) and oak forests (Bowersox and
McCormick 1987, Horsley 1988, Kolb et al. 1990). In recent
decades, hay-scented fern has become a large-scale forest
management problem in Pennsylvania; it has proliferated in
forest understories, disturbed areas, shelterwood cuts, and
clearcuts (Horsley 1981, Marquis et al. 1990, Groninger and
McCormick 1991, McWilliams et al. 1994).
Numerous mechanisms have been proposed to explain how
hay-scented fern interferes with the growth of hardwood seedlings. These include: reductions in photon flux density and

changes in light quality (R/FR ratio) under fern canopies
(Horsley 1986, Burnside 1988, Kolb et al. 1990, Horsley
1993a), competition for nutrients and water by fern root systems (Drew 1988), allelopathy (Horsley 1977, 1993a) and
physical occupation of organic soil horizons by fern root systems (Hammen 1993). Most research has focused on hayscented fern interference with growth of black cherry (Prunus
serotina Ehrh.) seedlings (Horsley 1993a, 1993b). There is
only limited information about mechanisms underlying fern
interference with other hardwood seedlings, including northern red oak (Quercus rubra L.) (Bowersox and McCormick
1987).
In recent decades, northern red oak has regenerated poorly
in parts of its range (Crow 1988). Although regeneration difficulties result from numerous interacting factors (Lorimer
1989, Kelly et al. 1990, Walters and Auchmoody 1993), interference by hay-scented fern has been cited as an important
factor limiting seedling growth and regeneration success in
some areas (Bowersox and McCormick 1987, Kolb et al. 1989,
Kolb and Steiner 1990, McWilliams et al. 1994).
In the current greenhouse study, reconstructed soil profiles
were used to approximate the soil types that red oak seedlings
typically face under field conditions. Red oak growth responses to subsoil liming and established fern covers were
examined. The greenhouse study was designed to eliminate the
effect of shading by the ferns and, under these conditions, test
the following hypotheses: (1) ferns reduce aboveground

growth of red oak; (2) ferns reduce the belowground growth of
red oak; (3) ferns alter the shoot/root ratio of red oak; (4) subsoil liming improves root growth and overall growth of red oak

924

LYON AND SHARPE

seedlings when grown alone or in the presence of ferns; and
(5) ferns reduce ectomycorrhizal infection frequency in red
oak.

Materials and methods
In June 1992, soil was collected from four field sites, Cookport, Clymer, Wharton, and Hazleton/Dekalb, in north-central
Pennsylvania. These sites support forests with at least half the
overstory composed of northern red oak (for detailed site
information, see Lyon and Sharpe 1995). All soil horizons to a
depth of 30 cm, but excluding the organic horizons, were
collected from excavated pits. Bulk densities of each mineral
horizon at the four field sites were determined. Each horizon
of the excavated soils was sieved through a 4-mm mesh.

Mineral soil horizons were then reconstructed in plastic, cylindrical growth cores (mini-terracosms; 15 cm diameter × 30 cm
depth). The growth cores had a total volume of 5025 cm3, a
volume/depth ratio of approximately 168, an adequate surface
area for O2 exchange, and ample water storage capacity (Hanson et al. 1987). Identical soil weights of individual horizons
(within each soil type) were placed in the growth cores and the
horizons were compacted to specific volumes to match measured field bulk density values (Reicosky et al. 1972). Calcium
amendments (subsoil liming) were achieved by inter-mixing
CaO at a rate of 500 kg Ca ha −1 in the deepest mineral horizon
of each soil type before horizon reconstruction. The rate of Ca
application was designed to reduce Al concentrations without
radically altering the balance of cations in the soil.
Organic horizons were collected in August 1992. Based on
soil analyses at each site, fern and non-fern samples were
chosen from areas of similar soil chemical status. Intact organic horizon cores (15.5 cm diameter × 3 cm depth) were used
to reduce soil disturbance and nitrification potential. Fern
fronds were cut back to the soil surface before cores were
removed from the ground. Each fern core was inspected to
ensure that all cores had similar rhizome densities and were
devoid of large non-fern roots or rocks. All recognized nonfern roots and large pieces of leaf litter were removed. Organic
horizons were immediately taken to the laboratory and

weighed. Cores used in the study were of the same weight
± 25 g. Small samples were collected from each core for
chemical analysis, and the cores were then fitted on top of the
appropriate mineral soil profiles. The completed mini-terracosms were wrapped with clear plastic and aluminum foil to
eliminate light penetration and reduce heat absorbency. The
bottom of each mini-terracosm was covered with a 400 µmmesh nylon screen equipped with a plastic cap connected with
a drainage tube to a collection bottle. The net volume of water
moving through the cores and the solution chemistry of the soil
effluent were determined. The mini-terracosms were kept in
the greenhouse and watered twice weekly with deionized
water until January 1993.
Soil analysis
Soil samples from each horizon (160 mini-terracosms × 3
horizons = 480 samples) were air-dried and sieved (2-mm

mesh) before analysis. Soil tests were conducted when the
soils were collected and at the end of the study. Exchangeable
cations were determined for a subsample (n = 2) from each
horizon (North Dakota State University 1988). Soil pH was
measured in 0.01 M CaCl2 (1/1, w/v). Extractable Ca and Al

were determined for all 480 samples in extracts of 0.01 M
SrCl2 (Joslin and Wolfe 1989). All SrCl2 soil extracts were
analyzed by atomic absorption spectrometry (Thermo-JarrelAsh (IL) Video 22 spectrometer, Milton, MA).
Red oak acorns and seedling growth
In October 1992, red oak acorns were collected from a single
parent tree growing on University Park Campus, Pennsylvania
State University, and immediately processed as described by
Olson (1974). After cold stratification, acorns of similar
weight (5.5 ± 0.1 g) were germinated on moist potting soil until
radicle emergence (1--2 cm long). Acorns were planted flush
with the surface of the organic horizons. Two weeks before
acorn planting, fern foliage was cut back to the soil surface to
ensure that adequate light was available for the red oak seedlings. This procedure simulated the phenology of fern growth
under field conditions, where early seedling growth generally
occurs before the emergence of fern fronds (Cody et al. 1977).
The impact of cutting on subsequent frond growth appeared to
be minimal; fronds grew back quickly and reached heights
similar to pre-cut levels (pre-cut grand mean height = 47.6 cm;
post-cut grand mean height = 49.7 cm). To prevent shading of
red oak seedlings during the study, fronds were restrained by

plastic rods and string. Red oak seedling cotyledons were
severed from all seedlings at Day 60 of the study to eliminate
their continued influence. The study was terminated 130 days
after acorn planting.
Greenhouse conditions and experimental design
Growth cores were placed on greenhouse benches equipped
with a 20% shade cloth. The greenhouse was illuminated
artificially to provide a 16-h photoperiod and temperature was
regulated by an automatic ventilation system. Maximum and
minimum temperatures ranged from 35 to 22 °C and from 30
to 14 °C, respectively. Relative humidity ranged from 30 to
90%. The experimental design was a completely randomized
block with replication. The experimental units were individual
mini-terracosms. There were four soil profiles, five treatments
and eight replications of each treatment (= 160 mini-terracosms). The five treatments were: (1) red oak grown alone; (2)
red oak grown with Ca-amended subsoil; (3) fern grown alone;
(4) fern grown with red oak; and (5) fern grown with red oak
with Ca-amended subsoil.
Starting in January 1993, each mini-terracosm was watered
twice a week with a simulated precipitation mix in an amount

proportional to the mean rainfall for the regions where the soils
were collected (2.6 cm per week). Volumes of artificial precipitation were increased by 20% after Day 42 to ensure an
adequate supply of water in the mini-terracosms. The composition of the simulated precipitation was modeled after the
1990 wet deposition chemistry records for Pennsylvania
(Lynch et al. 1991). A comparison of the composition and

FERN INTERFERENCE WITH RED OAK

properties of the simulated precipitation mix with 1990 ambient precipitation chemistry means are summarized in Table 1.
Sulfate concentrations in the artificial precipitation were
higher than ambient values to match the conductivity and pH
values.
Red oak and fern shoot and root growth analyses
Periodic measures of red oak height growth (every 7 to
10 days) were made. After 130 days, red oak seedlings and
ferns were harvested, red oak foliar and stem biomass were
determined, and mini-terracosms were separated by horizon.
Red oak and fern roots and rhizomes were separated by horizon, taking care to collect all fine roots. Plant tissues were then
washed in deionized water and frozen until processed. After
thawing, rhizomes and roots were carefully washed in dilute

Alconox® solution (Alconox, Inc., New York) followed by
three rinses in deionized water. Red oak root systems were then
separated into taproot and lateral roots, and fern tissues were
separated into rhizomes and fine roots. All plant tissues were
dried to constant weight at 80 °C to determine dry weight
biomass.
Measurements of red oak roots followed the terminology of
Lyford (1980). The primary laterals were counted in two categories (> 3 cm length and < 3 cm length). The total length of
the primaries > 3 cm long was also measured (Tennant 1975).
The five longest primary laterals within each horizon were
measured for diameter, total number of secondary laterals
> 2 cm long, and total length of secondary laterals. Total secondary lateral root length was calculated by multiplying the
combined length of the five longest secondary laterals by the
proportion of the remaining length of primary laterals. All
inter-horizon roots (roots not attached to the taproot) were
measured separately, but were added to primary length totals
for horizons in which they occurred. All root measurements
were analyzed by horizon but were summed across horizons to
calculate total root system measures. Mycorrhizal infection
frequency was determined as described by Mitchell et al.

(1984). Hay-scented fern rhizome measurements included total rhizome length, rhizome diameters of five randomly chosen

Table 1. Comparisons of the pH, specific conductance and chemical
composition of ambient versus artificial precipitation. All values are
means (± SE) and are based on eight sampling dates.

pH
Specific conductance (µs cm −1)
−1
SO 2−
4 (mg l )
NO −3 (mg l −1)
Cl − (mg l −1)
NH +4 (mg l −1)
K+ (mg l −1)
Na+ (mg l −1)
Ca 2+ (mg l −1)
Mg 2+ (mg l −1)

Ambient
precipitation

Artificial
precipitation

4.15
32.25
2.737
1.523
0.224
0.288
0.037
0.062
0.104
0.022

4.14 ± 0.015
32.58 ± 2.734
3.536 ± 0.024
1.409 ± 0.004
0.254 ± 0.007
0.312 ± 0.004
0.0289 ± 0.002
0.059 ± 0.019
0.059 ± 0.001
0.027 ± 0.001

925

rhizome segments, and the number, length and diameter of fine
roots found on each the five rhizome segments.
Statistical analysis
Statistical analysis was performed using Minitab 8.2 software
(Minitab, Inc. Data Tech Industries, Valley Forge, PA). Significance testing of continuous distribution variables between
treatments was conducted by ANOVA. Mean separations were
performed with Fisher’s protected LSD test. Comparisons between pre- and post-study soil parameters were made with
Dunnett’s test. All significant differences reported are at α =
0.05, unless otherwise noted. Variables were transformed when
necessary to meet the assumption of normality. Some measurements of red oak and fern belowground tissues were recalculated on a volumetric basis and others were normalized to
compare proportional allocation of roots by horizon. Linear
and polynomial regressions were also run using normalized
root morphological variables and root branching index values.
Differences in regression slope coefficients were determined
as described by Sokal and Rohlf (1981).

Results
During the 130-day study, there were no significant changes in
soil pH in the limed treatments of any of the four soils tested
(Table 2). Calcium concentrations in the limed treatments
exhibited significant changes in some soils; however, the Ca
concentrations of the limed horizons were still an order of
magnitude higher than those of the non-limed horizons at the
end of the study. Furthermore, Al concentrations at the end of
the study were lower in the limed soils than in the non-limed
soils. We conclude, therefore, that the effects of the Ca amendments were maintained throughout the study.
Fern biomass and biomass allocation
The only statistically significant soil type and treatment main
effect on fern biomass was on the fern shoot/belowground
biomass ratio (Table 3). However, the rhizome and total belowground biomass P-values were near significant. There were no
significant soil × treatment interactions on fern biomass, although treatment main effects were significant for both total
frond weight and fern shoot/belowground ratio. Frond weights
in the fern alone treatment were significantly higher than frond
weights in the red oak + fern treatment on all soils except the
Wharton soil, and they were significantly higher than frond
weights of ferns in the red oak + fern + Ca treatment across all
four soils (Table 3). Neither total belowground nor fine root
biomass was significantly different between treatments across
soil types (but note P-values). Fern shoot/belowground
biomass ratios reflected the reduction in shoot biomass relative
to constant values of belowground biomass. Fern shoot/belowground ratios were significantly lower in ferns in the red oak +
fern + Ca treatment than in ferns in either the fern alone or red
oak + fern treatments on the Cookport and Wharton soils. In
the Clymer and Hazleton/Dekalb soils, significant changes in
fern biomass allocation occurred only in response to the combined impacts of red oak and Ca.

926

LYON AND SHARPE

Table 2. Comparison of pH and 0.01M SrCl2-extractable Al and Ca in the lowermost mineral horizon B across soil types and treatments. Treatment
means of the soil parameters determined at the end of the study are compared to pre-experimental means using Dunnett’s test. Numbers in columns
followed by an asterisk are significantly different from the pre-experimental control value.
Treatment

B Horizon

B + Lime Horizon

pH

Al

Ca

pH

Al

Ca

Cookport
Pre-experimental values
Oak
Fern
Oak + fern

3.87
4.05
4.13*
4.06*

0.162
0.212*
0.201*
0.205*

0.026
0.019*
0.031
0.028

4.38
4.36

0.010
0.030*

0.634
0.470*

4.51

0.043*

0.755*

Clymer
Pre-experimental values
Oak
Fern
Oak + fern

4.17
4.16
4.11
4.14

0.387
0.121*
0.123*
0.183*

0.034
0.033
0.012*
0.020*

4.86
4.54

0.032
0.017

0.682
0.456

4.67

0.011

0.578

Wharton
Pre-experimental values
Oak
Fern
Oak + fern

3.64
3.88
3.83
3.82

0.176
0.301*
0.287*
0.267*

0.027
0.025
0.022
0.031

4.02
4.05

0.028
0.051*

0.735
0.306*

4.13

0.068*

0.593*

Hazleton/Dekalb
Pre-experimental values
Oak
Fern
Oak + fern

3.92
3.88
3.98
3.95

0.307
0.283
0.283
0.304

0.035
0.039
0.028*
0.026*

4.12
4.08

0.111
0.134

0.769
0.590*

4.15

0.122

0.516*

Table 3. Hay-scented fern biomass allocation (mean dry weights ± SE) at the end of the study.
Treatment

Total frond
(g)

Rhizomes
(g)

Fine roots
(g)

Total
belowground (g)

Shoot/belowground ratio

Cookport
Fern
Oak + fern
Oak + fern + Ca

6.74 ± 0.40
5.06 ± 0.45
4.30 ± 0.54

3.55 ± 0.40
2.60 ± 0.27
3.80 ± 0.55

1.55 ± 0.20
1.32 ± 0.14
1.42 ± 0.32

5.10 ± 0.56
3.92 ± 0.38
5.22 ± 0.83

1.32 ± 0.16
1.29 ± 0.19
0.82 ± 0.24

Clymer
Fern
Oak + fern
Oak + fern + Ca

6.92 ± 0.44
3.81 ± 0.36
4.37 ± 0.49

4.80 ± 0.45
3.90 ± 0.39
4.19 ± 0.38

1.98 ± 0.24
1.76 ± 0.17
1.35 ± 0.12

6.78 ± 0.50
5.41 ± 0.41
5.54 ± 0.37

1.02 ± 0.08
0.70 ± 0.10
0.79 ± 0.09

Wharton
Fern
Oak + fern
Oak + fern + Ca

5.48 ± 0.38
4.58 ± 0.36
4.03 ± 0.41

3.48 ± 0.37
2.71 ± 0.46
4.45 ± 0.49

1.53 ± 0.19
1.35 ± 0.14
1.44 ± 0.09

5.01 ± 0.42
4.06 ± 0.52
5.89 ± 0.55

1.09 ± 0.14
1.13 ± 0.19
0.68 ± 0.14

Hazleton/Dekalb
Fern
Oak + fern
Oak + fern + Ca

6.95 ± 0.47
4.93 ± 0.59
4.57 ± 0.42

3.15 ± 0.51
3.02 ± 0.47
3.97 ± 0.57

1.54 ± 0.26
1.70 ± 0.17
1.90 ± 0.24

4.69 ± 0.74
5.62 ± 0.57
5.87 ± 0.76

1.48 ± 0.39
0.88 ± 0.07
0.78 ± 0.18

Two-way ANOVA
Factor

P-Values
0.052
0.052
0.209

0.193
0.676
0.330

0.080
0.092
0.196

0.039
0.001
0.093

Soil
Treatment
Soil × Treatment

0.120
< 0.001
0.279

FERN INTERFERENCE WITH RED OAK

927

Figure 1 shows the vertical distribution of fern belowground
biomass by soil horizon volume (data were pooled across soils
because there were no significant soil or treatment main effects
or soil × treatment interactions). The shallow rooting pattern of
ferns in the growth cores (Figure 1) closely resembles the
growth form in the field (Hammen 1993). Rhizome biomass
accounted for 81--89% of the biomass in the organic horizons
and for 60--81% of the biomass in the entire soil profile.
Overall, the belowground biomass of ferns was not strongly
influenced by the red oak or subsoil Ca treatments.
Fern impacts on red oak growth
There were significant soil type (P = 0.038) and treatment
main effects (P < 0.001) on the height growth of red oak
seedlings (Figure 2). Mean final heights of red oak seedlings
in the red oak, red oak + Ca, red oak + fern and red oak + fern
+ Ca treatments were 45.83, 51.01, 19.86 and 28.23 cm,
respectively. On all four soils, final red oak height growth was
significantly reduced in the red oak + fern treatment compared
with the red oak and red oak + Ca treatments (Table 4).
Comparisons between red oak + fern and red oak + fern + Ca
treatments showed that there was significantly greater height
growth of red oak in the presence of subsoil Ca on the Wharton
and Hazleton/Dekalb soils, indicating that subsoil liming of
these soils counteracted the depression of red oak height
growth associated with the presence of ferns.
Significant soil type and treatment main effects were found
for red oak foliar, stem, total shoot, lateral root, taproot and
total root biomass (Table 4). Statistically significant soil ×
treatment interactions were found only for stem (P = 0.027)
and lateral root biomass (P = 0.026). For all measured variables, treatment main effect F-ratios were larger than either
soil main effect or soil × treatment interaction F-ratios, indicating that ferns exerted the dominant influence on red oak
biomass. Overall, seedlings in the red oak and red oak + Ca
treatments had significantly greater foliar, stem, total shoot,

Figure 1. Vertical distribution of the belowground biomass of hayscented fern weighted by soil horizon volume. Values presented are
means (± SE) and were calculated using pooled data from all four soil
profiles. Density (g per 1000 cm3) of belowground biomass was
calculated as total dry weight of rhizomes + fine roots.

Figure 2. Comparison of mean height growth of red oak seedlings on
four soils and across four treatments. Soil types are as follows: Cook
= Cookport; Clym = Clymer; Whar = Wharton; and Haz/D = Hazleton/Dekalb. Significant soil and treatment main effects and a significant soil × treatment interaction were found.

taproot and lateral root biomass than seedlings in the red oak
+ fern and red oak + fern + Ca treatments. This pattern was
evident within each soil type (Table 4).
Fern impacts on red oak shoot/root ratios
Shoot/root biomass ratios of red oak seedlings ranged from
0.81 to 1.93, with most at or near 1 (mean = 1.39) (Figure 3).
On the Clymer and Wharton soils, there were no significant
treatment effects on shoot/root ratios; however, on the Cookport and Hazleton/Dekalb soils, red oak seedlings allocated a
higher proportion of biomass to shoots than roots in the absence of ferns than in the presence of ferns.
Fern impacts on red oak root architecture
None of the treatments had a significant effect on red oak
taproot length in any of the four soils tested (Table 5). No
consistent patterns of root biomass reallocation were observed
despite the significant reduction in root biomass in the red oak
+ fern and red oak + fern + Ca treatments. The anticipated shift
in red oak root biomass deeper into the soil profile in response
to subsoil liming or the presence of hay-scented ferns only
occurred in the Cookport and Hazleton/Dekalb soils. Total
primary and secondary lateral lengths were significantly reduced in the presence of ferns, except for secondary laterals in
the red oak + fern treatment in Cookport soil (Table 5). Significant increases in root length in red oak + Ca treatments were
noted for primary laterals in Clymer soil and for secondary
laterals in Wharton soil. A significant decrease in secondary
laterals was noted in the red oak + fern + Ca treatment in
Cookport soil.
A graphical summary of root branching patterns of red oak
(i.e., total length of secondary laterals/total length of primary
laterals) in the organic horizons of the four soils is presented
in Figure 4. Significant treatment (P = 0.004) and soil effects
(P = 0.037) were noted. Among the four soils examined,
overall frequency of root branching was lowest in the Hazle-

928

LYON AND SHARPE

Table 4. Comparison of northern red oak seedling biomass (dry weight in g) allocation determined at the end of the 130-day growth study. Values
are means ± SE.
Soil and Treatment

Foliage

Stem

Total shoot

Lateral roots

Taproot

Cookport
Oak
Oak + Ca
Oak + fern
Oak + fern +Ca

6.8 ± 0.78
8.3 ± 0.78
2.2 ± 0.59
3.3 ± 0.83

5.6 ± 0.61
5.6 ± 0.87
1.6 ± 0.57
2.4 ± 0.57

12.4 ± 1.3
13.9 ± 1.5
3.8 ± 1.1
5.6 ± 1.4

2.96 ± 0.36
2.21 ± 0.24
0.90 ± 0.15
1.21 ± 0.21

5.92 ± 0.67
4.98 ± 0.79
2.71 ± 0.62
5.16 ± 0.52

8.88 ± 0.89
7.19 ± 0.95
3.61 ± 0.70
6.36 ± 0.71

Clymer
Oak
Oak + Ca
Oak + fern
Oak + fern + Ca

5.4 ± 0.50
5.5 ± 0.72
2.6 ± 0.55
2.8 ± 0.86

4.0 ± 0.41
4.5 ± 0.44
1.9 ± 0.65
1.8 ± 0.86

9.4 ± 0.8
10.0 ± 1.1
4.4 ± 1.2
4.6 ± 1.7

2.66 ± 0.31
3.10 ± 0.36
0.75 ± 0.12
0.77 ± 0.21

8.76 ± 0.99
7.14 ± 0.97
2.95 ± 0.47
4.96 ± 0.46

11.42 ± 1.05
10.24 ± 1.30
3.70 ± 0.44
5.73 ± 0.47

Wharton
Oak
Oak + Ca
Oak + fern
Oak + fern + Ca

5.3 ± 0.65
5.7 ± 0.79
2.3 ± 0.21
3.3 ± 0.49

4.0 ± 0.44
4.1 ± 0.80
1.4 ± 0.18
2.1 ± 0.33

9.3 ± 1.0
9.8 ± 1.6
3.7 ± 0.3
5.4 ± 0.8

1.70 ± 0.31
1.41 ± 0.43
0.29 ± 0.06
0.48 ± 0.10

6.79 ± 0.59
4.87 ± 0.81
2.64 ± 0.39
3.95 ± 0.67

8.49 ± 0.87
6.28 ± 0.87
2.92 ± 0.34
4.43 ± 0.76

Hazelton/Dekalb
Oak
Oak + Ca
Oak + fern
Oak + fern + Ca

6.6 ± 0.48
6.8 ± 0.72
3.0 ± 0.72
3.4 ± 0.54

4.7 ± 0.64
5.0 ± 0.21
2.0 ± 0.62
2.5 ± 0.46

11.1 ± 1.0
11.8 ± 0.6
5.0 ± 1.3
5.9 ± 0.9

1.69 ± 0.32
1.69 ± 0.25
0.59 ± 0.12
0.72 ± 0.19

5.72 ± 0.86
5.93 ± 0.71
4.12 ± 0.40
4.65 ± 0.45

7.41 ± 0.96
7.62 ± 0.67
4.71 ± 0.50
5.37 ± 0.63

Two-Way ANOVA
Factor

P-values

Soil
Treatment
Soil × Treatment

0.010
< 0.001
0.155

0.021
< 0.001
0.027

0.011
< 0.001
0.059

< 0.001
< 0.001
0.026

0.028
< 0.001
0.132

Total root

0.006
< 0.001
0.135

on Day 130. Each point in the regression represents a treatment
mean for a mineral soil horizon. Only mineral horizons were
used in the regressions because of difficulties in accurately
determining Al concentrations in the organic horizons. There
was a significant (P < 0.001) negative relationship between red
oak root branching and soil Al concentration (r 2 = 0.774).
Mycorrhizal infection frequency in red oak seedlings

Figure 3. Comparison of red oak seedling shoot/root biomass ratios at
the end of the study. Each bar is the mean (± SE) value for each soil ×
treatment combination.

ton/Dekalb soil. Root branching was significantly reduced in
the presence of ferns in all but the Wharton soil. Thus, with the
exception of the Wharton soil, inhibition of red oak root
branching was a major morphological and architectural effect
of fern interference.
Soil aluminum impacts on red oak root growth
Figure 5 illustrates the root branching response of red oak
seedlings to soil Al concentrations (0.01 M SrCl2 extractable)

Across treatments, there was a significant effect of soil
(P < 0.001) on the proportion of red oak primary laterals infected with ectomycorrhizal fungi, with the Wharton soil
showing the lowest overall infection frequency (mean =
3.98%) (Figure 6). There was also a significant (P < 0.001)
treatment effect on mycorrhizal infection frequency. Seedlings
in the red oak and red oak + Ca treatments had significantly
higher infection frequencies than seedlings in the red oak +
fern and red oak + fern + Ca treatments. On both the Cookport
and Wharton soils, mycorrhizal infection frequencies were
significantly reduced in the presence of ferns (red oak + fern
and red oak + fern + Ca treatments). No significant differences
in infection frequency across treatments were observed in the
Clymer soil. On the Hazleton/Dekalb soil, the highest infection frequency was observed when red oak was grown alone.

FERN INTERFERENCE WITH RED OAK

929

Table 5. Comparison of red oak seedling total taproot length, total primary lateral length, and total secondary lateral length between treatments.
Values are means ± SE.
Treatment
Taproot length (cm)
Oak
Oak + Ca
Oak + fern
Oak + fern + Ca
Total primary lateral length (cm)
Oak
Oak + Ca
Oak + fern
Oak + fern + Ca
Total secondary lateral length (cm)
Oak
Oak + Ca
Oak + fern
Oak + fern + Ca

Cookport

38.8 ± 2.68
36.7 ± 2.95
34.9 ± 1.13
39.5 ± 2.35

Clymer

41.4 ± 2.48
47.2 ± 2.45
48.1 ± 2.21
39.3 ± 1.28

Wharton

26.5 ± 1.49
32.0 ± 1.88
25.6 ± 1.86
25.3 ± 1.90

Hazleton/Dekalb

38.0 ± 2.35
43.8 ± 2.22
43.4 ± 2.16
41.5 ± 1.92

841 ± 33.3
620 ± 50.1
460 ± 45.9
451 ± 35.4

480 ± 41.8
587 ± 46.2
369 ± 29.1
384 ± 45.1

674 ± 34.2
589 ± 38.9
330 ± 36.9
335 ± 30.2

699 ± 41.1
695 ± 45.4
426 ± 24.4
451 ± 31.0

2028 ± 298
1661 ± 208
1401 ± 204
1036 ± 238

1522 ± 268
1852 ± 261
699 ± 155
614 ± 60

2737 ± 203
4082 ± 301
1552 ± 128
1509 ± 197

429 ± 43.2
456 ± 45.5
197 ± 31.7
279 ± 26.8

Figure 4. Root branching index (total length of secondary laterals/total
length of primary laterals) for northern red oak roots growing in the
organic horizons of the four soils. Mean branching index values (± SE)
are shown for each soil and treatment.

Discussion
To avoid light competition by herbaceous ground covers, intermediate shade-tolerant tree seedlings such as northern red oak
(Baker 1949) must achieve sufficient height growth to overtop
the canopy of ground vegetation. Because of its recurrent
growth form (Hanson et al. 1986), red oak has the potential for
multiple growth flushes in the same season, although typically
only a single flush each year is observed in the field (Crow
1988). In the present study, the first growth flush of red oak
seedlings showed no significant differences in height growth
or leaf area across soil types or treatments (Lyon 1995), suggesting that cotyledon reserves influenced early red oak
growth (Carpenter and Guard 1954, Hanson 1986, Crow
1988). However, cotyledons were removed at Day 60, and so
did not impact later phases of growth. The relatively robust

Figure 5. Second-order polynomial regression between 0.01 M SrCl 2extractable Al in the mineral soil horizons and log root branching
index (total secondary lateral length/total primary lateral length) for
red oak seedlings grown alone (s) or with fern (d). Each point
represents a treatment mean for mineral horizon soils only. The second-order regression was highly significant (P < 0.001; r 2 = 0.774).
Two horizons were excluded from this analysis because of contamination of the samples during processing.

height growth observed in seedlings in the fern-free treatments
was greater than height growth reported in native soils (McClenahen 1987, Joslin and Wolfe 1989, Kolb et al. 1990, Sharpe
et al. 1993). Our results may be explained by several factors,
including favorable growing conditions (light, humidity, temperature, soil water) and the existence of a multi-horizon
rooting environment with a nutrient-rich organic horizon. The
poor height growth exhibited by red oak seedlings in the
presence of ferns, despite the favorable growth conditions,
highlights the strong belowground interference by ferns.
Interspecific root competition between two plant species
occurs if one species (or both) exhibits a growth reduction

930

LYON AND SHARPE

Figure 6. Comparison of mean ectomycorrhizal infection rates (± SE)
in red oak seedlings across soil types and treatments.

when it shares the same soil resources as the other species, and
is not limited in growth by aboveground factors. In our study,
hay-scented fern interfered with the growth of red oak seedlings, despite the apparent lack of aboveground interference.
Red oak height growth and foliar and stem biomass were all
significantly reduced when seedlings were grown with ferns.
Similarly, Hippensteel (1992) noted that frond removal did not
restore height growth of first-year paper birch (Betula papyrifera Marsh.) or white ash (Fraxinus americana L.) seedlings to
that of non-fern treatments. Beck (1970) reported slow growth
rates of red oak seedlings released from overstory and understory competition in North Carolina. In contrast to our results,
Horsley (1993a) reported good height growth of black cherry
when fronds of competing ferns were restrained to allow full
light to reach the cherry leaf surfaces. The higher growth rate
of black cherry compared with red oak (Farmer 1980) may
account for the discrepancy between these species.
Many plant growth models predict a greater partitioning of
biomass to plant tissues responsible for capturing the most
limiting nutrient (Johnson and Thornley 1987). Several red oak
seedling studies have documented shifts in biomass partitioning to root systems in response to grass competition (Watson
1988, Kolb and Steiner 1990) and increasing irradiance (Bazzaz and Miao 1993). In our study, however, significant increases in biomass allocation to roots in response to ferns were
observed on only two of the four soils tested (Cookport and
Hazleton/Dekalb). Although these soil-specific patterns indicate the importance of soil type in predicting the biomass
allocation responses of red oak, they provide no information
on root architectural changes that might have occurred. Because assessment of root architectural changes can provide a
more comprehensive view of root responses to both root competition and the soil chemical environment than simple
biomass determinations (Fitter et al. 1991), we also determined
the effects of ferns on root branching patterns of red oak
seedlings.
We observed that red oak root branching (i.e., secondary
lateral growth) in the organic horizon was significantly reduced in the presence of hay-scented fern in all but the Wharton soil. This architectural change in red oak roots, which was
not reflected in the biomass allocation analysis, represents a
specific symptom of root interference between hay-scented

fern and red oak and has important implications for red oak
seedling growth because most root hairs in red oak are located
on secondary lateral roots (Richardson 1953). It is possible that
the reduction in fine root branching was a result of the physical
occupation of the organic horizons by fern rhizomes and fine
roots (McConnaughay and Bazzaz 1992). If ferns can interfere
with fine root development of red oaks in the organic horizons,
which typically contain higher concentrations of nutrients and
roots than mineral horizons (Vogt et al. 1983, Gale and Grigal
1987, Kelly and Joslin 1989, Ehrenfeld et al. 1992), they can
effectively diminish the availability of nutrients to red oaks.
The presence of Al in the subsoil also influenced red oak
seedling root growth and architecture. A significant negative
relationship between red oak root branching and 0.01 M SrCl2extractable Al was observed in both the fern and non-fern
treatments. This result is consistent with several previous oak
studies reporting that 0.01 M SrCl2-extractable Al was significantly and negatively correlated with red oak root branching in
mineral soil horizons (Kelly et al. 1990). We observed increases in red oak root branching in response to subsoil liming,
suggesting that soil Al contributed to poor red oak growth. It
has been widely recognized that first-year red oak seedlings
develop vertical taproots that are able to penetrate to a depth of
45 cm in the first year of growth (Korstian 1927, Rudolf 1929,
Toumey 1929, Holch 1931, Biswell 1935, Duncan 1941,
Richardson 1953, Carpenter and Guard 1954, Farmer 1975,
Lyford 1980). However, if high concentrations of Al in the
subsoil reduce fine root branching, then the ability of red oak
to exploit subsoil nutrients or water, or both, will be impaired.
Hay-scented fern also inhibited ectomycorrhizal infection in
red oak roots. Infection rates were relatively low in the current
study (4--38%) compared to other studies of red oak (Ruehle
1980, Mitchell et al. 1984). The low infection rates observed
on the Wharton soils are consistent with the finding that this
soil had the highest nutrient availability of the soils tested.
Although we did not measure fern mycorrhizal infection rates,
hay-scented fern has also been shown to have mycorrhizal
associations (Conard 1908).
We conclude that both hay-scented fern and subsoil Al
interfered with aboveground growth or belowground growth or
both of first-year northern red oak seedlings. Interference by
ferns was more severe than that caused by subsoil Al. The
presence of ferns resulted in reduced red oak height, shoot
growth, root growth, root branching in the organic horizon and
ectomycorrhizal infection rates. The degree of interference by
hay-scented fern was often soil specific. Liming treatments
were of limited effectiveness in improving red oak biomass in
the presence and absence of ferns. However, liming did result
in significant increases in red oak height growth in the presence of ferns on two of the four soils. Liming also ameliorated
the negative impacts of subsoil Al on red oak root branching.
Based on the observed changes in red oak root branching in
response to both ferns and subsoil Al, we conclude that
changes in root architecture may be a better indicator of belowground responses to ferns and subsoil Al than simple biomass
determinations.

FERN INTERFERENCE WITH RED OAK
Acknowledgments
Funding and logistical support for this study came from the School of
Forest Resources and the Environmental Resource Research Institute
at The Pennsylvania State University. We thank the Pennsylvania
Bureau of Forestry for access to field sites. Mary Kay Amistadi was
indispensable in the soil analyses. We thank Stephen Horsley for his
critical review of this manuscript, which greatly improved the quality
of this paper. Thanks are also extended to Dave DeWalle, Larry
McCormick, Todd Bowersox, and Dale Baker for their help in the
study.
References
Baker, F.S. 1949. A revised tolerance table. J. For. 47:179--181.
Bazzaz, F.A. and S.L. Miao. 1993. Successional status, seed size, and
responses of tree seedlings to CO2, light, and nutrients. Ecology
74:104--112.
Beck, D.E. 1970. Effect of competition on survival and height growth
of red oak seedlings. USDA For. Serv., Res. Pap. SE-56, 6 p.
Biswell, H.H. 1935. Effects of environment upon the root habits of
certain deciduous forest trees. Bot. Gaz. 96:676--708.
Bowersox, T.W. and L.H. McCormick. 1987. Herbaceous communities reduce the juvenile growth of northern red oak, white ash,
yellow poplar, but not white pine. Proc. 6th Central Hardwood
Forest Conf., pp 39--43.
Brach, A.R., S.J. McNaughton and D.J. Raynal. 1993. Photosynthetic
adaptability of two fern species of a northern hardwood forest. Am.
Fern J. 83:47--53.
Burnside, E.V. 1988. The effects of light intensity on the growth of
hay-scented fern. M.Sc. Thesis, The Pennsylvania State Univ., 54 p.
Carpenter, I.W. and A.T. Guard. 1954. Anatomy and morphology of
the seedling roots of four species of the genus Quercus. J. For.
52:269--274.
Cody, W.S., I.V. Hall and C.W. Cromptom. 1977. The biology of
Canadian weeds: 26. Dennstaedtia punctilobula (Michx.) Moore.
Can. J. Plant Sci. 57:1159--1168.
Conard, H.S. 1908. The structure and life history of the hay-scented
fern. Carnegie Inst. Wash. Publ. 94:1--106.
Crow, T.R. 1988. Reproductive mode and mechanisms for self-replacement of northern red oak (Quercus rubra)----a review. For. Sci.
34:19--40.
Drew, A.P. 1988. Interference of black cherry by ground flora of the
Allegheny uplands. Can J. For. Res. 18:652--656.
Drew, A.P. 1990. Fern and aster effects on black cherry shelterwood
regeneration. Can. J. For. Res. 20:1513--1515.
Duncan, W.H. 1941. A study of root development in three soil types in
the Duke Forest. Ecol. Monogr. 11:142--164.
Ehrenfeld, J.G., E. Kaldor and R.W. Parmelee. 1992. Vertical distribution of roots along a toposequence in the New Jersey pinelands.
Can. J. For. Res. 22:1929--1936.
Farmer, R.E. 1975. Dormancy and root regeneration of northern red
oak. Can J. For. Res. 5:176--185.
Farmer, R.E. 1980. Comparative analysis of 1st-year growth in six
deciduous tree species. Can. J. For. Res. 10:35--41.
Fitter, A.H., T.R. Strickland, M.L. Harvey and G.W. Wilson. 1991.
Architectural analysis of plant root systems. 1. Architectural correlates of exploitation efficiency. New Phytol. 118:375--383.
Gale, M.R. and D.F. Grigal. 1987. Vertical root distribution of northern
tree species in relation to successional status. Can J. For. Res.
17:829--834.
Groninger, J.W. and L.H. McCormick. 1991. Effects of sulfometuron
on hay-scented fern spore emergence. Can. J. For. Res. 21:942--943.

931

Hammen, S.C.L. 1993. Density-dependent phenotypic variation in the
hay-scented fern, Dennstaedtia punctilobula. Bull. Torrey Bot.
Club. 120:392--396.
Hanson, P.J. 1986. Studies of Quercus rubra L. seedling dry matter
accumulation, morphological development, and carbon dioxide
exchange under controlled conditions. Ph.D. Thesis, Univ. Minnesota, 169 p.
Hanson, P.J., R.E. Dickson, J.G. Isebrands, T.R. Crow and R.K.
Dixon. 1986. A morphological index of Quercus seedling ontogeny for use in studies of physiology and growth. Tree Physiol.
2:273--281.
Hanson, P.J., R.K. Dixon and R.E. Dickson. 1987. Effects of container
size and shape on the growth of northern red oak seedlings. HortScience 22:1293--1295.
Hippensteel, T.E. 1992. Light stress induced by hayscented fern.
M.Sc. Thesis, School of Forest Resources, The Pennsylvania State
Univ., 87 p.
Holch, A.E. 1931. Development of roots and shoots of certain deciduous tree seedlings in different forest sites. Ecology 12:259--298.
Horsley, S.B. 1977. Allelopathic inhibition of black cherry by fern,
grass, goldenrod, and aster. Can. J. For. Res. 7:205--216.
Horsley, S.B. 1981. Control of herbaceous weeds in Allegheny hardwood forests with herbicides. Weed Sci. 29:655--662.
Horsley, S.B. 1984. Hayscented fern rhizome development in uncut
and thinned Allegheny hardwood stands. Am. J. Bot. 71:80--81.
Horsley, S.B. 1986. Evaluation of hayscented fern influence with
black cherry. Am. J. Bot. 73:668--669.
Horsley, S.B. 1988. How vegetation can influence regeneration. In
Proc. Guidelines for regenerating Appalachian Hardwood Stands.
Eds. H.C. Smith, A.W. Perkins and W.E Kidd. SAF Publ., 88-03,
West Virginia Univ. Books, pp 38--55.
Horsley S.B. 1993a. Mechanisms of interference between hay-scented
fern and black cherry. Can. J. For. Res. 23:2059--2069.
Horsley, S.B. 1993b. Evaluation of allelopathic interference between
hay-scented fern and black cherry. J. Chem. Ecol. 19:2737--2755.
Horsley, S.B. and D.A. Marquis. 1983. Interference by weeds and deer
with Allegheny hardwood reproduction. Can. J. For. Res. 13:61--69.
Johnson, I.R. and J.H.M. Thornley. 1987. A model of shoot:root
partitioning with optimal growth. Ann. Bot. 60:133--142.
Joslin, J.D. and M.H. Wolfe. 1989. Aluminum effects on northern red
oak seedling growth in six forest soil horizons. Soil Sci. Soc. Am.
J. 53:274--281.
Kelly, J.M. and J.D. Joslin. 1989. Mass and chemical composition of
roots in two second-growth oak forests in eastern Tennessee. For.
Ecol. Manage. 27:87--92.
Kelly, J.M., M. Schaedle, F.C. Thornton and J.D. Joslin. 1990. Sensitivity of tree seedlings to aluminum: II Red oak, sugar maple, and
European beech. J. Environ. Qual. 19:172--179.
Kolb, T.E., T.W. Bowersox, L.H. McCormick and K.C. Steiner. 1989.
Effects of shade and herbaceous vegetation on first-year germination and growth of direct-seeded northern red oak, white ash, white
pine, and yellow poplar. In Proc. 7th Central Hardwood For. Conf.
Eds. G. Rink and C.A. Budelsky. USDA For. Serv. Gen. Tech. Rep.
NC-132, pp 156--161.
Kolb, T.E. and K.C. Steiner. 1990. Growth and biomass partitioning
response of northern red oak genotypes to shading and grass competition. For. Sci. 36:293--303.
Kolb, T.E., T.W. Bowersox and L.H. McCormick. 1990. Influences of
light intensity on weed-induced stresses of tree seedlings. Can. J.
For. Res. 20:503--507.
Korstian, C.F. 1927. Factors controlling germination and early survival in oaks. Yale School For. Bull. No. 19, CT, 61 p.

932

LYON AND SHARPE

Lorimer, C.G. 1989. The oak regeneration problem: new evidence on
causes and possible solutions. In Proc. 17th Ann. Symp. Hardwood
Res. Council, Merrimac, WA, pp 23--40.
Lyford, W.H. 1980. Development of the root system of northern red
oak (Quercus rubra L.) Harvard For. Pap. 21, MA, 14 p.
Lynch, J.A., J.W. Grimm and E.S. Corbett. 1991. Atmospheric deposition: spatial and temporal variations in Pennsylvania----1990. ERRI,
Penn. State Univ., Publ. No. ER9103, 79 p.
Lyon, J. 1995. The impacts of soils, herbaceous competition and soil
liming on the growth and nutrition of northern red oak (Quercus
rubra L.) seedlings. Ph.D. Thesis, Intercollege Program in Ecology,
The Pennsylvania State Univ., 169 p.
Lyon, J. and W.E. Sharpe. 1995. Impacts of electric deer exclusion
fencing and soils on plant species abundance, richness, and diversity following clearcutting in Pennsylvania. In Proc. 10th Central
Hardwood Forest Conf. Eds. K.W. Gottschalk and S.L.C. Fosbroke.
USDA For. Serv. Gen. Tech. Rep. NE-197, pp 47--59.
Maquire, D.A. and R.T.T. Forman. 1983. Herb cover effects on tree
seedling patterns in a mature hemlock-hardwood forest. Ecology
64:393--401.
Marquis, D.A., R.L. Ernst and S.L. Stout. 1990. Prescribing silvicultural treatments in hardwood stands of the Alleghenies. USDA For.
Serv. Gen Tech. Rep. NE-96, 32 p.
McClenahen, J.R. 1987. Effects of simulated throughfall pH and
sulfate concentration on a deciduous forest soil. Water Air Soil
Pollut. 35:319--333.
McConnaughay, K.D.M. and F.A. Bazzaz. 1992. The occupation and
fragmentation of space: consequences of neighboring roots. Funct.
Ecol. 6:704--710.
McWilliams, W.H., S.L. Stout, T.W. Bowersox and L.H. McCormick.
1994. Advance tree-seedling regeneration and herbaceous cover in
Pennsylvania’s forests. Penn. For. 94:14--17.
Mitchell, R.J., G.S. Cox, R.K. Dixon, H.E. Garrett and I.L. Sander.
1984. Inoculation of three Quercus species with eleven isolates of
ectomycorrhizal fungi. II. Foliar nutrient content and isolate effectiveness. For. Sci. 30:563--572.

Olson, D.F., Jr. 1974. Quercus L. Oak. In Seeds of Woody Plants in the
United States. USDA Agric. Handbook 450:692--703.
Reicosky, D.C., R.J. Millington and D.B. Peters. 1972. Patterns of
water uptake and root distribution of soybeans (Glycine max) in the
presence of a water table. Agron. J. 64:292--297.
Richardson, S.D. 1953. A note on some differences in root hair
formation between seedlings of sycamore and American oak. New
Phytol. 52:80--82.
Rudolf, P.O. 1929. A study of the seedling root systems of certain
forest trees. M.F. Thesis, Cornell Univ., 72 p.
Ruehle, J.L. 1980. Ectomycorrhizal colonization of container-grown
northern red oak as affected by fertility USDA For. Serv. Res. Note
SE-297, 7 p.
Sharpe, W.E., B.R. Swistock and D.R. DeWalle 1993. A greenhouse
study of northern red oak seedlings growth on two forest soils at
different stages of acidification. Water Air Soil Pollut. 66:121--133.
Sokal, R.P. and F.J. Rohlf. 1981. Biometry. W.H. Freeman and Company, New York, 859 p.
Tennant, D. 1975. A test of a modified line intersect method of
estimating root length. J. Ecol. 63:995--1001.
Toumey, J.W. 1929. Initial root habit in American trees and its bearing
on regeneration. Proc. 4th Intl. Bot. Congr. 1:713--725.
Vogt, K.A., E.E. Moore, D.J. Vogt, M.J. Redlin and R.L. Edmonds.
1983. Conifer fine root and mycorrhizal biomass within the forest
of Douglas-fir stands of different ages and site productivities. Can.
J. For. Res. 13:429--437.
Walters, R.S. and L.R. Auchmoody. 1993. Factors limiting northern
red oak reproduction in Pennsylvania. USDA For, Serv. GTR NC161, pp 271--280.
Watson, G.W. 1988. Organic mulch and grass competition influence
tree root development. J. Aboric. 14:200--203.