Forest Ecology and Management 303 (2013) 81–90
Contents lists available at SciVerse ScienceDirect
Forest Ecology and Management
journal homepage: www.elsevier.com/locate/foreco
Review
A meta-analysis of the effect of forest management for timber
on understory plant species diversity in temperate forests
Marlyse C. Duguid ⇑, Mark S. Ashton
Yale School of Forestry & Environmental Studies, 360 Prospect St., New Haven, CT 06511, USA
a r t i c l e
i n f o
Article history:
Received 29 November 2012
Received in revised form 2 April 2013
Accepted 4 April 2013
Keywords:
Clearcut
Harvesting
Herbs
Logging
Selection systems
Silviculture
a b s t r a c t
Many studies have examined affects of forest management—particularly regeneration treatments—for
timber on understory plant diversity. These studies taken independently show no clear trends in diversity
with degree and/or periodicity of disturbance from timber harvests. Here we present a meta-analysis
synthesizing primary field research on response of understory plant diversity to timber harvesting in
temperate forests, particularly in North America. Across a pool of 96 studies, we find no effect on understory plant species richness from managing forests for timber. When intensive regeneration harvests (e.g.
clearcut, shelterwood) are separated from less intensive regeneration harvests (e.g. single tree and group
selection systems) and thinnings, selection harvests show a positive effect on species richness. Intensive
regeneration harvests and thinning treatments had no effect on species richness. We examined the role of
stand development following regeneration treatments, and found no detectable effects on species richness for even-aged stands within the first 50 years after clearcut and shelterwood timber harvests. Stands
in later successional stages, however, had lower species richness than un-logged stands. All these findings
together suggest that silvicultural activities focused toward timber management are not inconsistent
with conservation of understory plant diversity. We suggest site-specific characteristics (e.g. resource
availability, resource heterogeneity) at various temporal and spatial scales, have a larger role to play in
defining understory plant diversity than the disturbance of harvesting itself. Managers therefore should
consider underlying factors of site and species composition, and should examine regionally specific studies when planning silvicultural treatments. In addition, it should be noted that our analysis makes no distinction in classifying the nature of diversity, especially between colonizing early-successional species
that peak after 1–10 years and then disappear, and late successional, often more site specific and shade
tolerant species, that may persist post harvest but often disappear or retract in their range and abundance. Further studies are needed to tease out differences in diversity in relation to successional stage
and affects of forest management.
Ó 2013 Elsevier B.V. All rights reserved.
Contents
1.
2.
3.
4.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.
Data selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2.
Data classification and calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.
Structure of data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.
Meta-analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1.
General impacts to understory diversity from timber harvesting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.2.
Comparison in understory diversity between even- and uneven-aged silviculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.3.
Changes in understory diversity with even-aged stand development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4.4.
Challenges of applying meta-analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
⇑ Corresponding author. Address: Yale University School of Forestry and Environmental Studies, Marsh Hall, 360 Prospect St., New Haven, CT 06511, USA. Tel.: +1
203 650 9118.
E-mail address: Marlyse.duguid@yale.edu (M.C. Duguid).
0378-1127/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.foreco.2013.04.009
82
82
82
83
85
85
85
85
85
86
86
87
82
M.C. Duguid, M.S. Ashton / Forest Ecology and Management 303 (2013) 81–90
4.5.
Recommendations for reporting research and data archiving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Management implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix A.
Supplementary material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.
1. Introduction
Forests account for roughly a third of global land cover, and are
home to much of the planet’s biodiversity (FAO, 2010; UNEP,
2010). Biologically diverse systems, they also serve a variety of human needs. Of these needs, timber and pulp are of critical importance, and the effects of their extraction are the focus of a large
body of research. While often overlooked, much of the biodiversity
in forests as well as many non-timber forest products and other
ecosystem services are provided by understory plants (Whigham,
2004). Although these species are not specifically targeted in timber extraction, harvest activities are known to affect understory
plant communities (Roberts and Gilliam, 2003). The degree to
which harvest intensity, site, and successional process interact to
drive these changes is less certain.
Many researchers have investigated the effects of even-aged
harvesting on understory plant diversity with mixed results. A review by Roberts and Gilliam (2003) examining clearcut harvesting
in eastern North America found no clear pattern predicting understory plant diversity response. Moola and Vasseur (2008) conducted a similar review, but focused on late-successional forests
types of northeast North America, they found only small effects
of clearcutting on understory plant species richness. Observed increases in understory plant diversity soon after (<20 years) clearcut harvesting may be a result of early successional colonizers
(Jenkins and Parker, 1999; Battles et al., 2001; Brosofske et al.,
2001; Moola and Vasseur, 2004; Kreyling et al., 2008; Loya and
Jules, 2008). Although this spike in richness is not always observed
(Meier et al., 1995; Nagaike et al., 1999; Scherer et al., 2000). There
is evidence that understory residual plant diversity generally declines after clearcutting of late successional forests (Moola and
Vasseur, 2008). Studies examining thinning treatments also show
inconsistent effects on understory plant species richness. Some
show positive effects on diversity (Thomas et al., 1999; Metlen
and Fiedler, 2006), others negative (Wyatt and Silman, 2010),
and some no effect (Wayman and North, 2007; Schwilk et al.,
2009). Uneven-aged harvesting through selection regeneration
methods has also failed to show clear trends. Falk et al. (2008)
found increased understory plant species richness in both singletree and group selection treatments, while Jenkins and Parker
(1999) found increases in group-selection, but decreases in single-tree treatments. Other studies have found single-tree selection
to have either positive effects (Scheller and Mladenoff, 2002) or no
effect (Kern et al., 2006). Many of these results may depend on
what groups of plants were being investigated (residual, colonizing, total), the condition and successional age of the forest, and
the forest type examined. Successional stage is an important consideration, particularly following even-aged silvicultural regeneration methods (clearcut, seed tree, shelterwood) (Smith et al., 1997).
The structure and amount of competition and resources available
on the forest floor changes more dramatically, and phases of development move through time more uniformly (initiation, stem
exclusion, understory reinitiation, old growth) as compared to uneven-aged methods (Oliver and Larson, 1996). Therefore understory plant diversity is expected to be more temporally dynamic
in even-aged regeneration methods.
In summary, there has been a considerable amount of research
focused on the response of understory plant diversity following
88
88
88
89
89
forest management, but there has not been a systematic analysis
of all these studies taken together. Of particular importance is
the role of harvest treatment intensity in mediating effects on
understory plants. Harvesting treatments vary greatly depending
on whether the goal is to promote regeneration (clear cut, seed
tree, shelterwood), promote growth of the existing stand (thinning) or a combination (selection). Past studies have focused on
specific forest types, specific regions, or particular forest conditions
(i.e. old growth). Through meta-analysis we seek to identify patterns over the complete temperate forest biome, including both
coniferous and broadleaved forests. In this paper we integrate current research to answer the following questions. Firstly, does forest
management through timber harvesting have a negative effect on
understory plant diversity in temperate forest systems? Secondly,
does the type of silvicultural treatment (i) even-aged regeneration
methods (clear cutting/shelterwoods), (ii) thinning, (iii) uneven
aged regeneration methods (selection) define those effects? Lastly,
are successional patterns of understory plant species diversity
apparent following even-aged methods of regeneration?
2. Methods
2.1. Data selection
We performed a meta-analysis examining understory plant species diversity and timber harvesting in temperate forests, particularly focused in North America. We chose to analyze species
richness because it is both the simplest and the most commonly
reported diversity measure (Magurran, 2004). Also in many circumstances it can be easily extrapolated from datasets on understory plant communities, even when not directly reported,
allowing us to include more data.
We performed an intensive literature search through three databases (Google Scholar, Scirus, and ISI Web of Science) with the
keywords: forest, understory, diversity, and logging – and then
completed additional searches first substituting harvesting for logging, and then richness for diversity. When necessary, values were
extracted from figures using the program Data Thief (Tummers,
2006). In circumstances where richness data was collected, but
not reported, or where no measure of variance or sample size
was disclosed every effort was made to acquire those data by contacting primary authors. In experimental designs that tested more
than one variable (e.g. thinning and burning) only the control and
the silvicultural harvesting treatment were analyzed. Comparative
studies that looked at more than one silvicultural treatment (e.g.
clearcut vs. thinning) were analyzed separately, as were studies
that examined more than one group of plants. We included
control-treatment comparisons, and allowed studies utilizing
observational chronosequences, but excluded diachronic studies
that lacked a true control (i.e. before-after comparisons only).
Control conditions for comparisons included both ‘‘old growth’’
and ‘‘mature’’ stands as defined by individual researchers. These
controls encompassed a diverse range of land-use and disturbance
histories (see Supplementary materials). In studies that reported
repeated measures only final values were used. We insured that
results were only used once when reported in more than one
paper. We included studies that examined understory response,
83
M.C. Duguid, M.S. Ashton / Forest Ecology and Management 303 (2013) 81–90
but the definition of ‘‘understory’’ varied amongst publications.
Most studies used a height limitation (e.g. all vascular
plants 6 1 m), although others used only herbaceous species or
specific growth forms. All studies examined response of vascular
plant diversity in the ground-story stratum at minimum, but three
included bryophytes.
or uneven-aged management (selection). Based on the Peet and
Christensen (1988) and Oliver and Larson (1996) models of forest
succession we further divided the ‘‘even-aged’’ studies into three
categories; (1) initiation (0–15 years since harvest), (2) stem exclusion (16–50 years since harvest), (3) later successional phases
(>50 years since harvest). We included studies described as ‘‘intensively logged’’ or ‘‘unregulated harvest’’ in the even-aged category.
For each comparison we collected mean species richness, standard
deviation, and sample size for both control and experimental treatments. We used these data to calculate an effect size for each comparison; in this analysis, the effect of forest harvesting on
understory plant species richness. Therefore a positive effect size
indicates an increase in species richness. We used the standardized
2.2. Data classification and calculations
We set up a database of logged vs. unlogged comparisons from
37 papers published between 1992 and 2010 (Table 1). We categorized the data by silvicultural treatment, even-aged regeneration
methods (clearcut, seed tree, shelterwood); thinning treatments;
Table 1
Summary table of studies used for this meta-analysis. For each comparison we list the source publication (study), sample size (n), sample mean (x), and standard deviation (s) for
each treatment. Control is indicated by a subscript c, and treatment with a subscript t. Treatment type (Tr) is indicated as even-aged (EA), selection (S), or thinning (TH).
Additionally, even-aged studies are classified by stand phase, initiation (I), 0–15 years since harvest; stem exclusion (SE), 16–50 years since harvest; and later successional phases
(LS), >50 years since harvest. Two measure of effect size are given, (d) the standardized mean difference using hedges’ adjusted g and its corresponding variance (vd), and the log
transformed ratio of means (L) and its corresponding variance (vL).
Study
nc
nt
xc
xt
sc
st
Tr
Phase
Battles et al. (2001)
Brosofske et al. (2001)
Bryce (2009)
Halpern et al. (2005)
Halpern et al. (2005)
Halpern et al. (2005)
Halpern et al. (2005)
Jenkins and Parker (1999)
Jenkins and Parker (1999)
Kreyling et al. (2008)
Loya and Jules (2008)
Meier et al. (1995)
Moola and Vasseur (2004)
Moola and Vasseur (2004)
Nagaike et al. (1999)
Roberts and Zhu (2002)
Scherer et al. (2000)
Wayman and North (2007)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Bryce (2009)
Bryce (2009)
Bryce (2009)
Gilliam (2002)
Jenkins and Parker (1999)
Jenkins and Parker (1999)
Kern et al. (2006)
Kreyling et al. (2008)
Loya and Jules (2008)
Moola and Vasseur (2004)
Ramovs and Roberts (2003)
Selmants and Knight (2003)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Yorks and Dabydeen (1999)
Burton et al. (2009)
D’Amato et al. (2009)
Duffy and Meier (1992)
Duffy and Meier (1992)
Duffy and Meier (1992)
Duffy and Meier (1992)
6
16
20
60
192
192
149
12
12
16
19
15
9
9
5
42
4
3
28
28
28
28
28
28
28
28
28
28
28
28
28
28
20
20
20
30
12
12
3
16
19
9
6
8
28
28
28
4
16
20
19
20
19
4
15
41
132
191
192
72
8
11
16
19
15
5
5
17
64
4
3
28
28
28
28
28
28
28
28
28
28
28
28
28
28
17
18
30
30
11
7
3
16
16
3
6
11
28
28
28
8
8
20
19
20
19
7.7
26.6
28.4
15.15
15.93
15.93
16.46
27
26.6
19.63
14.35
13.9
8.44
8.44
45.2
13.44
17
3.57
5.4
5.4
5.4
5.4
5.4
5.4
5.4
10.3
10.3
10.3
10.3
5.7
5.7
5.7
28.4
28.4
28.4
13.2
27
26.6
24.4
19.63
14.35
8.44
9.4
18.66
10.3
5.7
5.7
26
13.56
11.2
9
9.95
14.53
15.6
33
89.4
13.36
14.54
18.01
13.71
28.4
37.9
23.63
33.95
10.1
10
9.93
41.8
13.44
14.5
2.83
4.9
5
6.7
7
4.3
9.1
6.9
5.4
5.9
5.6
3.4
8.8
6.4
5.8
31.6
40.8
52.7
13.9
29.9
39
25.2
25.81
22.19
6.56
7.3
26.17
5.2
5.1
4.6
33.5
6.88
7.25
7.35
8.75
6.04
2.4
5.5
7.16
7.35
7.51
7.51
5
6.58
6.24
4.39
7.21
2.96
1.8
1.8
13.3
3.11
3.54
1.84
2.12
2.12
2.12
2.12
2.12
2.12
2.12
3.17
3.17
3.17
3.17
2.12
2.12
2.12
7.16
7.16
7.16
2.74
6.58
6.24
4.89
4.39
7.21
1.8
0.2
2.55
3.17
2.12
2.12
4.24
4.44
4.18
2.65
1.87
4.13
1.8
3.5
34.58
6.85
6.23
6.58
4.17
5.94
10.61
4.46
9.09
1.94
1.97
4.76
10.2
1.92
2.87
1.04
1.59
2.12
2.12
2.12
1.59
3.17
2.12
2.12
1.59
2.12
1.59
4.23
2.65
2.12
18.55
13.58
42.72
3.83
5.97
11.11
5.57
6.88
13.69
0.19
0.2
1.89
2.65
2.12
2.12
10.24
2.35
3.06
2.59
3.93
3.64
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
SE
SE
SE
SE
SE
SE
SE
SE
SE
SE
SE
SE
SE
SE
SE
LS
LS
LS
LS
LS
LS
d
3.25
1.34
2.09
0.25
0.20
0.29
0.58
0.21
1.27
0.88
2.34
1.48
0.79
0.45
0.30
0.00
0.67
0.40
0.26
0.19
0.60
0.74
0.58
1.35
0.70
1.79
1.73
1.72
2.71
0.91
0.29
0.05
0.23
1.14
0.71
0.21
0.44
1.43
0.12
1.04
0.72
1.08
9.69
3.28
1.72
0.28
0.51
0.78
1.65
1.06
0.62
0.38
2.14
vd
L
vL
0.945
0.158
0.110
0.024
0.010
0.011
0.021
0.209
0.209
0.137
0.177
0.170
0.333
0.318
0.261
0.039
0.528
0.680
0.072
0.072
0.075
0.076
0.074
0.088
0.076
0.100
0.098
0.71
0.22
1.15
0.13
0.09
0.12
0.18
0.05
0.35
0.19
0.86
0.32
0.17
0.16
0.08
0.00
0.16
0.23
0.10
0.08
0.22
0.26
0.23
0.52
0.25
0.65
0.56
0.098
1.11
0.43
0.12
0.02
0.11
0.36
0.62
0.05
0.10
0.38
0.03
0.27
0.44
0.25
0.25
0.34
0.68
0.11
0.21
0.25
0.68
0.43
0.20
0.13
0.88
0.020
0.003
0.007
0.006
0.002
0.002
0.002
0.010
0.012
0.005
0.017
0.005
0.013
0.051
0.021
0.002
0.021
0.134
0.009
0.012
0.009
0.009
0.010
0.010
0.009
0.009
0.006
0.61
0.011
0.013
0.011
0.010
0.023
0.009
0.025
0.004
0.009
0.016
0.030
0.008
0.037
0.005
<.001
0.003
0.013
0.011
0.013
0.018
0.021
0.016
0.011
0.012
0.023
0.137
0.079
0.072
0.071
0.110
0.123
0.088
0.067
0.179
0.280
0.668
0.142
0.123
0.493
4.245
0.499
0.098
0.072
0.074
0.400
0.245
0.114
0.110
0.102
0.165
(continued on next page)
84
M.C. Duguid, M.S. Ashton / Forest Ecology and Management 303 (2013) 81–90
Table 1 (continued)
Study
nc
nt
xc
xt
sc
st
Tr
Phase
d
vd
L
vL
Duffy and Meier (1992)
Duffy and Meier (1992)
Duffy and Meier (1992)
Duffy and Meier (1992)
Duffy and Meier (1992)
Jackson et al. (2009)
Loya and Jules (2008)
Moola and Vasseur (2004)
Scheller and Mladenoff (2002)
Wyatt and Silman (2010)
Bryce (2009)
Bryce (2009)
Bryce (2009)
Bryce (2009)
Bryce (2009)
Falk et al. (2008)
Falk et al. (2008)
Jenkins and Parker (1999)
Jenkins and Parker (1999)
Jenkins and Parker (1999)
Jenkins and Parker (1999)
Jenkins and Parker (1999)
Jenkins and Parker (1999)
Jenkins and Parker (1999)
Jenkins and Parker (1999)
Kern et al. (2006)
Kern et al. (2006)
Kern et al. (2006)
Kern et al. (2006)
Scheller and Mladenoff (2002)
Converse et al. (2006) (data from Schwilk et al. (2009))
Dodson and Peterson (2010)
Harrod et al. (2007) (data from Schwilk et al. (2009))
Metlen and Fiedler (2006)
Outcalt (2005) (data from Schwilk et al. (2009))
Phillips and Waldrop (2008) (data from Schwilk et al. (2009))
Ritchie (2005) (data from Schwilk et al. (2009))
Stephens and Moghaddas (2005) (data from Schwilk et al., 2009)
Takafumi and Hiura (2009)
Waldrop et al. (2008) (data from Schwilk et al. (2009))
Waldrop et al. (2008) (data from Schwilk et al. (2009))
Wayman and North (2007)
Youngblood et al. (2006) (data from Schwilk et al. (2009))
20
18
10
20
20
13
19
9
4
180
20
20
20
20
20
90
90
12
12
12
12
12
12
12
12
3
3
3
3
4
4
4
4
3
3
2
3
3
16
3
6
3
4
20
18
10
20
20
13
25
3
4
180
35
37
33
33
31
90
224
7
13
11
13
2
15
7
21
3
3
3
3
4
4
4
4
3
3
3
3
3
19
3
3
3
4
10.4
11.6
11.36
10.65
9.55
78.5
14.35
8.44
4.27
9.8
28.4
28.4
28.4
28.4
28.4
4.3
4.3
27
27
26.6
26.6
27
27
26.6
26.6
24.4
24.4
24.4
24.4
4.27
3.33
28.25
6.545
57.3
1.913
1.59
1.393
4.33
57.94
3.02
1.943
3.57
23.93
7.3
4.94
2.5
7.45
7.55
78.6
24.54
4.72
6.51
7.4
72.2
55.8
51.9
70.2
59.5
3.6
3.7
21.9
22.9
30.4
29.5
35
33.3
42.9
40.1
25.67
28.93
26.27
26.13
10.64
4.3
26.71
6.358
66.2
2.313
2.68
1.76
5.75
56.56
2.94
1.56
4.22
21.6
3.31
2.62
5.01
2.44
2.56
15.8
7.21
1.8
0.93
0.2
7.16
7.16
7.16
7.16
7.16
2.85
2.85
6.58
6.58
6.24
6.24
6.58
6.58
6.24
6.24
4.89
4.89
4.89
4.89
0.93
1.37
4.79
0.96
4.68
0.59
0.16
0.4
1.9
5.36
0.48
0.59
1.84
3.86
3.12
3.1
3
1.94
2.25
15.79
13.38
1.18
2.51
0.2
46.74
46.84
48.83
20.68
2.78
1.9
1.5
7.14
6.49
3.98
6.13
11.31
7.75
6.09
8.71
3.32
0.5
6.82
3
2.56
1.76
2.61
0.406
7.1
1.927
0.936
0.631
0.69
5.77
0.6
0.148
1.5
6.084
EA
EA
EA
EA
EA
EA
EA
EA
EA
EA
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
TH
TH
TH
TH
TH
TH
TH
TH
TH
TH
TH
TH
TH
LS
LS
LS
LS
LS
LS
LS
LS
LS
LS
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
0.94
2.27
2.05
1.42
0.81
0.01
0.90
2.03
1.03
11.97
1.15
0.71
0.59
2.43
6.17
0.29
0.30
0.72
0.61
0.69
0.45
1.06
0.84
2.52
1.66
0.24
1.04
0.25
0.34
2.87
0.53
0.35
0.22
1.18
0.22
1.02
0.56
0.79
0.24
0.12
0.67
0.31
0.40
0.111
0.183
0.306
0.125
0.108
0.154
0.102
0.615
0.566
0.210
0.091
0.081
0.084
0.136
0.456
0.022
0.016
0.240
0.168
0.185
0.164
0.623
0.163
0.393
0.173
0.672
0.757
0.672
0.676
1.016
0.518
0.508
0.503
0.783
0.671
0.937
0.693
0.719
0.116
0.668
0.525
0.675
0.510
0.35
0.85
1.51
0.36
0.23
0.00
0.54
0.58
0.42
0.28
0.93
0.68
0.60
0.90
0.74
0.18
0.15
0.21
0.16
0.13
0.10
0.26
0.21
0.48
0.41
0.05
0.17
0.07
0.07
0.91
0.26
0.06
0.03
0.14
0.19
0.52
0.24
0.28
0.02
0.03
0.22
0.17
0.10
0.014
0.025
0.163
0.006
0.008
0.006
0.025
0.026
0.049
<.001
0.015
0.022
0.030
0.006
0.003
0.008
0.006
0.020
0.011
0.006
0.008
0.057
0.009
0.007
0.007
0.019
0.013
0.036
0.018
0.026
0.084
0.010
0.006
0.006
0.264
0.046
0.070
0.069
0.001
0.022
0.018
0.131
0.026
mean difference (SMD) in species richness between logged forests
(experimental group, X E ) and unlogged forests (control group, X E )
for our effect size. The unbiased effect size (d) for each comparison
was calculated using the following equation:
d¼
"
#
XE XC
J
s
ð1Þ
where X E X C is the difference between the means of the experimental and control groups, and s is the pooled standard deviation
Eq. (2):
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðNE 1ÞðsE Þ2 þ ðNC 1ÞðsC Þ2
s¼
ðNE þ N C 2Þ
ð2Þ
and J is a bias correction factor Eq. (3):
J ¼1
3
4ðNE þ NC 2Þ 1
ð3Þ
where NE and NC are the sample sizes of the experimental and control groups respectively, and SE and SC are their standard deviations
(Hedges and Olkin, 1985).
We also calculated the log transformed response ratio (L) Eq.
(4):
L ¼ lnðX E Þ lnðX C Þ
and the standard error associated with L (SEL) Eq. (5):
ð4Þ
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
!
u
u
1
1
SEL ¼ ts
þ
n1 ðx1 Þ2 n2 ðx2 Þ2
ð5Þ
In this study the log-transformed response ratio represents the
percentage change between logged and unlogged stands. The response ratio describes the proportional change observed after harvesting, and are commonly used in ecology because they provide
more information on the magnitude of effects than SMD (Hedges
et al., 1999). Both analyses include pooled standard deviation.
Incorporating experimental variance into the effect size provides
more meaningful results.
We chose to apply random-effects models to calculate overall
effect size (d+). Random-effects models are commonly accepted
as the most appropriate methods for meta-analysis in ecology.
Studies included in this analysis are not identical in their methods
and site characteristics, and these differences introduce variability.
Random-effects models allow for variability of effect size amongst
studies and treats heterogeneity between studies as random. Additionally, under the random-effects model there is not one single
true effect size, but a range of possible effects (Borenstein et al.,
2009; Viechtbauer, 2010).
We created seven models to explore our research questions. The
first model pooled all comparisons to look at the overall effect of
logging on understory species richness. We then ran models
for each silvicultural treatment type (even-aged regeneration,
85
M.C. Duguid, M.S. Ashton / Forest Ecology and Management 303 (2013) 81–90
thinning, or selection), and within the even-aged treatment type
we examined successional stage—running separate models for
initiation, stem exclusion, and later successional stand phases. All
models utilized the DerSimonian–Laird estimate and inverse
weighting for pooling (Hedges, 1981; DerSimonian and Laird,
1986). We examined between study variation in each group using
a heterogeneity measure (Q), calculated by weighting the sum of
squared differences between individual effects and the pooled
effect, tested against a chi-square distribution (Hedges and Olkin,
1985). The null hypothesis is that variation between studies is zero,
therefore we considered the group heterogeneous when p(Q) 6 .05.
To check for publication bias we visually inspected a funnel plot of
effect size by standard error and found no indication of bias in the
data. All analyses were conducted using R version 2.15.2, including
the package metafor version 1.6 (Viechtbauer, 2010; R Development Core Team, 2012).
detected. Studies within each treatment category, however,
showed considerable variation with positive, negative, or no significant difference (Fig. 1).
Lastly, when we included the grouping factor of successional
stage within the even-aged regeneration methods we found stands
in later successional stages had lower species richness than unharvested controls (n = 16, SMD = 1.54, 95% CI 2.85 to 0.24), a
28.4% decrease (95% CI 41.18% to 15.52%) in understory species
richness from the controls. This decrease was not seen in the earlier development phases of stand initiation (n = 32, SMD = 0.10,
95% CI 0.22 to 0.41; L = 4.08%, 95% CI 9.14% to 17.31%) and stem
exclusion (n = 15, SMD = 0.20, 95% CI 0.36 to 0.77; L = 6.81%, 95%
CI 9.99% to 23.62%). Between study heterogeneity was high
(p < .0001) for all models except thinning (Table 2).
4. Discussion
3. Results
4.1. General impacts to understory diversity from timber harvesting
3.1. Structure of data
There is a widespread presumption that timber harvesting is a
major cause of biodiversity loss (Wood et al., 2000). Our analysis
suggests that the story is more complicated for understory plant
diversity. We sought to use meta-analysis to determine whether
timber harvesting, defined broadly, has a negative effect on understory plant diversity, across forests of the temperate biome, primarily within North America. Comparing harvested stands with
unharvested controls we found no clear influence on understory
plant species richness. We did find that the type of harvesting (silvicultural treatment) and successional stage do determine response. Our analysis supports the supposition of Roberts and
Gilliam (2003)—they refute making broad generalizations about
understory plant response. We agree with their call for closer
examination of different harvesting techniques and or forest types
(see Sections 4.2 and 4.3).
According to Halpern and Spies (1995), timber harvesting has
two major classes of effects on understory plant communities; (i)
initial effects, which affect individuals in the pre-disturbance community; and (ii) long-term effects on developing plant populations.
The mechanisms influencing understory plant community response within these classes vary (e.g. survival and dispersal for initial effects; succession and competition for long-term effects)
(Roberts and Gilliam, 2003). Understory plant communities in temperate forests are diverse and dynamic. Shorter generation times,
as well as greater environmental sensitivity and specificity mean
that the understory plant community responds faster and at finer
scales to disturbance and site variation (Ellum, 2007, 2009; Fahey
and Puettmann, 2007; Ellum et al., 2010). These factors make
understory plant communities ideal subjects for examining response to disturbance. Moreover, there are many site-specific factors influencing understory plant species diversity, most
importantly resource availability and heterogeneity (Bartels and
Chen, 2009).
Our dataset was comprised of 96 logged vs. unlogged comparisons from 37 papers. The number of comparisons from each paper
ranged from one to fourteen. The comparisons were largely evenaged (63 studies), with 13 thinning and 20 selection treatments
(Table 1). Within the even-aged category 32 comparisons were
from the initiation stage of stand development, 15 from stem
exclusion, and 16 from later successional stages. The overwhelming majority of studies represented North America with only two
studies outside representing Japan. Twenty-one of the studies
had a coniferous element with four from the Southern US, four
from the northeast, and the remaining from Western North
America.
3.2. Meta-analysis
Firstly, across all studies, irrespective of silvicultural treatment
(clearcut, shelterwood, selection, thinning) or successional stage,
timber harvesting had no clear influence on understory plant richness (n = 96, SMD = 0.02; 95% CI 0.24 to 0.29). The effect size
measure by the response ratio indicates a slight increase of 4.9%
(95% CI 1.4% to 11.27%) in understory species diversity under forest management as compared to unmanaged, but this increase is
not significant (Table 2 and Fig. 1).
Secondly, the only silvicultural treatment with a positive effect
on understory richness was selection (n = 20, SMD = 0.95; 95% CI
0.45–1.49)], with a 30% (95% CI 12.38–47.66%) increase in understory plant diversity as calculated by the response ratio. Both
even-aged regeneration methods (clearcut, shelterwood) (n = 63,
SMD = 0.29, 95% CI 0.64 to 0.05; L = 3.16%, 95% CI 9.91% to
3.59%) and thinning (n = 13, SMD = 0.03, 95% CI 0.34 to 0.40;
L = 1.8%, 95% CI 6.11% to 9.71%) had no significant effects
Table 2
Summary table of random effects meta-analysis models. SMD, standardized mean difference using Hedges’ adjusted g; 95% confidence intervals; n, number of comparisons; Q,
heterogeneity across studies; p(Q), heterogeneity tested against a chi-square distribution; % change (SE), difference in richness between control and treatment based calculated as
a log transformed ratio of means; ⁄, significant effect (p 6 .05).
Model
SMD
95% CI
n
Q
p(Q)
% Change
Overall
Even-aged
Initiation
Stem exclusion
Later successional⁄
Selection⁄
Thinning
0.02
0.29
0.10
0.20
1.54
0.97
0.03
(0.24, 0.29)
(0.64, 0.05)
(0.22, 0.41)
(0.36, 0.77)
(2.85, 0.24)
(0.45, 1.49)
(0.34, 0.40)
96
63
32
15
16
20
13
1522.01
1281.52
356.87
116.09
649.79
201.81
6.98
<0.0001
<0.0001
<0.0001
<0.0001
<0.0001
<0.0001
0.8592
4.9% (3.2)
3.2% (3.4)
4.1% (6.8)
6.8% (8.6)
28.4% (6.5)
30% (9.0)
1.8% (4.0)
86
M.C. Duguid, M.S. Ashton / Forest Ecology and Management 303 (2013) 81–90
Fig. 1. Forest plot of all individual comparisons used in this analysis. The standardized mean effect size (SMD) is represented by the mid-point of the box, and the line the 95%
confidence interval. Box sizes are drawn proportional to the precision of the estimates, and represent the weight given to the study in the model. The diamond below the
studies represents the overall effect in the random effects model. The studies have been separated by treatment type and stand phase; A, Even-aged/initiation; B, even-aged/
stem exclusion; C, even-aged/later successional; D, selection; E, thinning.
4.2. Comparison in understory diversity between even- and unevenaged silviculture
Timber harvesting can fall anywhere in a broad spectrum of disturbance intensities. We expected that even-aged silvicultural
treatments (such as clearcuts)—which have a higher single impact
disturbance intensity—would have more dramatic effects on
understory plant diversity than less intense management types,
such as single tree selection. Our results did indeed demonstrate
that the type of silvicultural treatment determines the magnitude
and direction of response for understory plant diversity. Surprisingly, selection systems—viewed by many as the least disruptive
silvicultural treatment (Quinby, 1991; Mladenoff and Pastor,
1993; Parker, 1993)—was the only silvicultural treatment with a
clear positive effect (30% increase) on understory plant species
richness.
Our analysis did not distinguish between single-tree and group
selection, but in some cases group selection treatments showed
higher positive effects than single-tree selection treatments (Jenkins and Parker, 1999). Greater resource availability and heterogeneity resulting from gap creation in group selection cuts may
account for some of these effects. The heterogeneity may provide
opportunities for early successional ruderals to co-exist within—
but spatially separate from—late successional species, presumably
restricted to less disturbed areas (Canham and Marks, 1985; Fahey
and Puettmann, 2007). If resource availability was the principal
consideration we would expect a positive response to more intense
even-aged regeneration treatments, which we did not observe. Indeed, Fredericksen et al. (1999) found no relationship between
residual basal area and understory diversity, supporting the argument that resource heterogeneity may be more important than
availability in disturbed forests (Bartels and Chen, 2009). Even-
aged treatments for timber harvesting have the potential to reduce
microsite heterogeneity. Additionally, the uniformity of these intense disturbance treatments could potentially shift community
composition from late successional toward ruderal, with no observed change in richness (Meier et al., 1995; Small and McCarthy,
2002). Furthermore, although silvicultural systems may have similar effects to natural disturbance in some contexts (Franklin et al.,
2002), it is important to examine them separately if the goal is to
promote biodiversity in managed landscapes.
4.3. Changes in understory diversity with even-aged stand
development
The main conceptual model of understory plant diversity following a stand-replacing disturbance is based on changes in competition for resources with a developing forest canopy (Peet and
Christensen, 1988; Roberts and Gilliam, 2003). It is thought that
understory plant diversity spikes when resources are abundant
immediately following the disturbance and drops after canopy closure (stem exclusion), during the period of intense competition
and self-thinning in the canopy, and high shade at ground level.
Over time, stem exclusion gives way to understory re-initiation;
resources in the understory slowly become more available as stand
stature and stratification increases in the forest canopy. Compositional change during old-growth may either result in a small drop
in species richness or reach a new peak (Fig. 2.) (Oliver and Larson,
1996; Roberts and Gilliam, 2003). Other models assume a constant
increase in herbaceous layer diversity through succession (Christensen and Gilliam, 2003). Of course these patterns will be influenced by disturbance intensity and spatial scale.
In contrast to the theoretical model of stand development just
described, our results indicate no difference in understory richness
M.C. Duguid, M.S. Ashton / Forest Ecology and Management 303 (2013) 81–90
Fig. 2. A graphical representation of the relationship between stand successional
phase and understory diversity. The solid black line represents Roberts and
Gilliam’s conceptual model of understory diversity through forest succession
(2003). Stand phases are as defined by Oliver and Larson (1996). Diversity is initially
high during initiation responding to an increase in available resources following a
disturbance, drops off with increasing overstory competition during stem exclusion,
and then slowly increases as the stand transitions towards later successional stages.
In the ‘‘Old Growth’’ stage richness may either reach a new peak because of
increased niche-specificity (a) or decline from the loss of early successional species
(b). The results of this meta-analysis are represented by the dotted line. No
detectable difference between treatment and controls in ‘‘stand initiation’’ and
‘‘stem exclusion’’ phases, lower diversity in the later successional phases. We are
hesitant to project into ‘‘old growth’’ as it is a qualitative definition, and our
analyses were based on stand age.
following even-aged silvicultural treatments (clearcut, shelterwood) in the first 50 years of stand development, when compared
with unharvested controls. Our results do suggest, however, that
later successional stages of forest development (>50 years) treated
by even-aged silviculture have lower diversity (a 28% decline) than
unharvested controls. Our results support Duffy and Meier’s (1992)
assertion that even-aged silviculture has long-lasting effects on
understory plant species diversity, but contradicts many studies
whose findings demonstrate no perceivable effects from clearcutting on understory plant diversity after several decades (Albert
and Barnes, 1987; Ruben et al., 1999; Scheller and Mladenoff, 2002).
The interaction between successional stage and disturbance
should also be considered. Bartels and Chen (2009) found resource
heterogeneity is particularly important for maintaining understory
plant diversity in disturbed and later successional forests. Intense
even-aged silvicultural harvesting methods may be lethal enough
to substitute early successional plants for late successional through
site scarification, use of fire and/or deposition of slash. Although
diversity between young even-aged stands and unlogged stands
may not be significant, species composition might be. We suggest
that as young even-aged stands mature they become more homogeneous in structure and more uniformly limiting in shade and
microhabitats, extinguishing early successional shade-intolerants,
and by inference lowering plant diversity over time (Bartels and
Chen, 2009). Rotation time should also be considered. The end of
a rotation coinciding with a period of depressed understory species
richness could create a feedback with the potential to degrade the
understory plant community. This could have long-term consequences for the understory composition—especially for residual
species associated with later successional stages, sensitive to disturbance and potentially dispersal limited (Whigham, 2004). Our
results suggest that the conclusions of Duffy and Meier (1992)
for the southern Appalachians—understory herb communities cannot recover in a single rotation cycle (40–150 years)—may hold
true in other systems. Further research aimed at monitoring understory plant communities through multiple rotations could investigate these long-term effects and further elucidate understory plant
diversity in managed landscapes.
Resource heterogeneity associated with the complex structure
of old growth stands should support the greatest diversity of species (Halpern and Spies, 1995). Our analysis does not follow this
pattern. There are two possible divergent explanations. One explanation is that the structures associated with late successional
87
unmanaged forests are not apparent in our definition of late successional (P50 years), which in successional time is still relatively
young. The second explanation is one we have previously discussed, namely understory diversity may never return to pre-harvest levels (Duffy and Meier, 1992; Meier et al., 1995). In the latter
case different guilds of plants are predominant at different successional stages (Grime, 1977; Hooper and Vitousek, 1997; Whigham,
2004), so stand phase will likely influence compositional diversity,
a metric not addressed in this analysis. Clearly further meta-analyses are needed to investigate such compositional changes, but for
this to occur more carefully documented studies are needed.
4.4. Challenges of applying meta-analysis
The value of meta-analysis lies in its ability to synthesize a
group of primary studies and statistically test the overall effect
without the subjective interpretation necessary in traditional reviews (Gurevitch et al., 1992). There are challenges when combining studies with disparate methodologies, from across an entire
biome. One confounding factor is the range of ecological conditions
(e.g. climate, soil, forest type) and ‘‘control’’ treatments (e.g. differing land-use history, age). Another is the various definitions of
‘‘understory’’ used across studies. We acknowledge that some level
of taxonomic grouping and the use of morphospecies is common,
that these methods are not consistent across studies, and that
these groupings may not always be disclosed. Additionally, the
nature of control-impact studies and observational chronosequences makes it impossible to insure that initial states for harvested and control stands are identical. The use of meta-analysis
allows us to test whether there is a clear effect of forest management on understory plant species diversity when examined over
a broad region—valuable regardless of these inter-study differences. Each individual study included in the analysis considered
underlying site conditions and scale and we examined a standardized response, therefore we did not additionally control for these
factors in our models.
The degree of heterogeneity can be quantified by the Q statistic
(Hedges et al., 1999), and is high in all but one of our models (Table 2). Some of this heterogeneity is addressed by the subdivision
of the data, and by applying random effects models. Unfortunately,
the number of studies available was not enough for subdivision
into all homogeneous models (p(Q) > .05), which would be ideal.
While the thinning group did have this characteristic, the group
was comprised largely of studies done in collaboration with the fire
surrogate study (Schwilk et al., 2009) therefore methodologies
were more similar than studies in the other groups we examined.
The degree of heterogeneity acceptable for these types of analyses
in experimental ecology is constantly under debate (Gurevitch
et al., 2001; Stewart, 2010). More importantly, the high variability
observed between studies is an important secondary result on its
own (Hedges et al., 1999). It supports the assertion that understory
plant communities should be examined within their site-specific
context. Although disturbance drives understory response, the
site-specific nature of plant communities and autecology of species
plays a large role in that response (Roberts and Gilliam, 2003;
Takafumi and Hiura, 2009). Although scale likely accounts for some
of the discrepancies between studies, the amount is unclear. For
example, Thomas et al. (1999) found an increase in understory
plant species richness in response to thinning across spatial scales,
while Metlen and Fiedler (2006) found increases at the plot level,
but no differences in richness at larger spatial scales. Thus, this
interaction between site conditions and spatial scale may also play
a role in the heterogenity between studies.
Another important consideration when looking at species richness is the issue of sampling intensity. To truly estimate species
richness in a population sample values should be extrapolated
88
M.C. Duguid, M.S. Ashton / Forest Ecology and Management 303 (2013) 81–90
using either parametric or non-parametric models (Magurran,
2004). Of all the studies we analyzed few used richness estimators,
and within that subset the models utilized varied. Finally, the most
evident constraint on our analysis is our dependence on species
richness as a proxy for diversity. Although species richness has a
long history of popularity and use, it is not informative about community composition. Furthermore, it is only one small component
of total ecosystem diversity (Noss, 1990; Magurran, 2004). It would
be more informative to incorporate a compositional component
into the analysis. Although most studies reported some compositional component in their results, original datasets for each experiment were not available, which would be necessary to apply
meta-analysis in that manner. Additionally, structural and functional diversity are likely more important than species diversity
for various ecosystem services (interception, infiltration, nutrient
uptake), and the two may be disparately affected following harvesting (Ernst et al., 2006). Each individual species in a community
has specific relationships with regards to ecosystem function, and
the use of plant functional groups may allow for more informative
ecological comparisons (Hooper and Vitousek, 1997; Tilman et al.,
1997; Diaz and Cabido, 2001). The use of functional groups is rife
with complication; plant functional types can include an extensive
array of species traits including both effect and response traits
(Lavorel and Garnier, 2002). Including a functional-trait analysis
in a meta-analytic context is additionally complicated by the variability of classification systems between scientists and regions. An
accepted standard for ecological grouping that could be applied
across a wide range of studies would aid in applying future
meta-analyses in this context. Despite all these limitations, studies
of understory plant richness continue to be important. Species
richness is an easily collected and understood metric; richness values each species individually, which is important because we do
not fully understand all the ecological roles of any individual species (Burton et al., 1992; Magurran, 2004).
4.5. Recommendations for reporting research and data archiving
With the increased use of quantitative data reviews in ecology,
and their potential to be used in informing policy, there are a number of ways that further research could be structured to facilitate
more meaningful analysis. Similar to Paillet et al. (2010) we found
many papers lacking summary statistics, we support their assertion that journals require reporting of variance and sample size.
In addition, a standard of reporting for site conditions would be
helpful; requiring a series of specific variables to be included with
Supplemental materials (e.g. soils, aspect, slope) for each dataset
would facilitate analysis of studies into more specific groupings.
Most importantly is the issue of data archiving. In recent years
many journals and funding agencies have established policies
requiring original data archiving (Whitlock, 2011). For our analyses, many papers had missing summary statistics, and when we requested this information the vast majority of authors were unable
to locate their original datasets. Of those that were could, many
were unable to access them because of software incompatibility.
We assert the framework proposed by Whitlock (2011) is a good
start, and that all journals should require complete species lists
and datasets included as Supplemental materials (Pärtel, 2006;
Paillet et al., 2010). Availability of these data would allow integration of compositional and functional-trait analyses even if those responses were not originally investigated.
5. Management implications
The role of resource availability, site-specificity, and species
composition all combine to define plant diversity, and therefore
community responses to a disturbance such as timber harvesting.
Moving forward research that seeks to draw larger conclusions
should include measures such as compositional or functional
diversity. In addition, replication of studies within identical forest
systems could be used to help quantify the scale in which site-differences affect plant diversity. Furthermore, our analyses highlight
the complications in drawing conclusions across the North American temperate biome. However, we summarize the following management implications:
1. General impacts from timber harvesting and silvicultural treatments are not necessarily inconsistent with maintaining understory plant diversity in temperate forested systems at the local
scale.
2. Uneven-aged selection systems show higher species richness
than unharvested controls, possibly because of greater temporal
and spatial resource heterogeneity allowing late-successional
and early successional species to co-exist. The compositional
component needs further examination.
3. Even-aged silvicultural practices within young stands (less than
50 years in age) show no differences in understory plant diversity as compared to unmanaged late successional forest. Older
stands (greater than 50 years in age) that originated from
even-aged silviculture, however, are lower in understory plant
diversity than unmanaged late successional stands. We suggest
that plant diversity declines with time in stands managed as
even-aged because early successional plants represent a larger
proportion of diversity and late successional species not present
in young stands and slow to colonize in older ones.
4. The choice of rotation length in even-aged systems is important,
with the potential to degrade understory species diversity.
Although the choice of rotation length is based on a number
of environmental and economic variables, we assert that understory diversity should be included as a consideration, especially
in areas with rare or endangered species.
5. Forest managers should be keenly aware of underlying site and
species composition, and should look to regionally specific
studies when planning silvicultural treatments.
Examining diversity at larger spatial scales would provide more
insight into the complexities of understory plant diversity, as the
results of this study are only applicable at the stand level. Alpha
diversity should be considered, but overall diversity may be higher
in a mosaic of lower richness patches, than a homogeneous landscape with higher alpha richness. Even-aged silviculture maintained within short rotations is necessary for successful
regeneration and management of many timber tree species, while
thinning and selection treatments fulfill different management
objectives (Smith et al., 1997). We argue that maintaining a diverse
silvicultural portfolio, including even-aged, uneven-aged, and hybrid models (e.g. irregular even-aged systems) within a managed
landscape is the best way to enhance diversity.
Acknowledgements
We gratefully acknowledge all of the researchers who contributed data for this analysis, especially those who had to dig deeply
into their archives. We would like to thank David Skelly, Kristofer
Covey, and two anonymous referees for their valuable comments
on previous versions of this manuscript. We are also grateful to
Thomas James, Kendra James, and Jeff Carroll for advice and support during the development of this paper. This research is supported by a fellowship to the senior author provided by The
School of Forestry and Environmental Studies, Yale University
and The New York Botanical Garden.
M.C. Duguid, M.S. Ashton / Forest Ecology and Management 303 (2013) 81–90
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.foreco.2013.04.009.
References
Albert, D.A., Barnes, B.V., 1987. Effects of clearcutting on the vegetation and soil of a
sugar maple-dominated ecosystem, western upper Michigan. Forest Ecology
and Management 18, 283–298.
Bartels, S.F., Chen, H.Y.H., 2009. Is understory plant species diversity driven by
resource quantity or resource heterogeneity? Ecology 91, 1931–1938.
Battles, J.J., Shlisky, A.J., Barrett, R.H., Heald, R.C., Allen-Diaz, B.H., 2001. The effects
of forest management on plant species diversity in a Sierran conifer forest.
Forest Ecology and Management 146, 211–222.
Borenstein, M., Hedges, L.V., Higgins, J.P.T., 2009. Introduction to Meta-Analysis.
Wiley, Hoboken, NJ, USA.
Brosofske, K.D., Chen, J., Crow, T.R., 2001. Understory vegetation and site factors:
implications for a managed Wisconsin landscape. Forest Ecology and
Management 146, 75–87.
Bryce, E., 2009. Influence of Silvicultural Treatment, Site Characteristics, and Land
Use History on Native and Nonnative Forest Understory Plant Composition on
the Penobscot Experimental Forest in Maine. M.S. Thesis, University of Maine,
Orono, ME. pp. 135.
Burton, P.J., Balisky, A.C., Coward, L.P., Cumming, S.G., Kneeshaw, D.D., 1992. The
value of managing for biodiversity. The Forestry Chronicle 68, 225–237.
Burton, J.I., Zenner, E.K., Frelich, L.E., Cornett, M.W., 2009. Patterns of plant
community structure within and among primary and second-growth northern
hardwood forest stands. Forest Ecology and Management 258, 2556–2568.
Canham, C.D., Marks, P.L., 1985. The response of woody plants to disturbance. In:
Pickett, S.T.A., White, P.S. (Eds.), The Ecology of Natural Disturbance and Patch
Dynamics. Academic Press, Orlando, FL, p. 472.
Christensen, N.L., Gilliam, F.S., 2003. Temporal and spatial patterns of herbaceous
layer communities in the North Carolina Piedmont. In: Gilliam, F.S., Roberts,
M.R. (Eds.), The Herbaceous Layer in Forests of Eastern North America. Oxford,
New York, pp. 224–237.
Converse, S.J., Block, W.M., White, G.C., 2006. Small mammal population and habitat
responses to forest thinning and prescribed fire. Forest Ecology and
Management 228, 263–273.
DerSimonian, R., Laird, N., 1986. Meta-analysis in clinical trials. Controlled Clinical
Trials 7, 177–188.
D’Amato, A.W., Orwig, D.A., Foster, D.R., 2009. Understory vegetation in old-growth
and second-growth Tsuga canadensis forests in western Massachusetts. Forest
Ecology and Management 257, 1043–1052.
Diaz, S., Cabido, M., 2001. Vive la différence: plant functional diversity matters to
ecosystem processes. Trends in Ecology and Evolution 16, 646–655.
Dodson, E.K., Peterson, D.W., 2010. Dry coniferous forest restoration and understory
plant diversity: The importance of community heterogeneity and the scale of
observation. Forest Ecology and Management 260, 1702–1707.
Duffy, D.C., Meier, A.J., 1992. Do appalachian herbaceous understories ever recover
from clearcutting? Conservation Biology 6, 196–201.
Ellum, D., 2007. Demographic patterns and disturbance responses of understroey
vegetation in a managed forest of southern New England: implications for
sustainable forestry and biodiversity. In: Forestry and Environmental Studies.
Yale, New Haven, p. 235.
Ellum, D.S., 2009. Floristic diversity in managed forests: demography and
physiology of understory plants following disturbance in southern New
England forests. Journal of Sustainable Forestry 28, 132–151.
Ellum, D.S., Ashton, M.S., Siccama, T.G., 2010. Spatial pattern in herb diversity and
abundance of second growth mixed deciduous-evergreen forest of southern
New England, USA. Forest Ecology and Management 259, 1416–1426.
Ernst, R., Linsenmair, K.E., Rödel, M.-O., 2006. Diversity erosion beyond the species
level: dramatic loss of functional diversity after selective logging in two tropical
amphibian communities. Biological Conservation 133, 143–155.
Fahey, R.T., Puettmann, K.J., 2007. Ground-layer disturbance and initial conditions
influence gap partitioning of understorey vegetation. Journal of Ecology 95,
1098–1109.
Falk, K.J., Burke, D.M., Elliott, K.A., Holmes, S.B., 2008. Effects of single-tree and
group selection harvesting on the diversity and abundance of spring forest
herbs in deciduous forests in southwestern Ontario. Forest Ecology and
Management 255, 2486–2494.
FAO, 2010. Global forest resources assessment. In: Nations, Food and Agriculture
Organization of the United Nations, Rome, pp. 1–375.
Franklin, J.F., Spies, T.A., Pelt, R.V., Carey, A.B., Thornburgh, D.A., Berg, D.R.,
Lindenmayer, D.B., Harmon, M.E., Keeton, W.S., Shaw, D.C., Bible, K., Chen, J.,
2002. Disturbances and structural development of natural forest ecosystems
with silvicultural implications, using Douglas-fir forests as an example. Forest
Ecology and Management 155, 399–423.
Fredericksen, T.S., Ross, B.D., Hoffman, W., Morrison, M.L., Beyea, J., Johnson, B.N.,
Lester, M.B., Ross, E., 1999. Short-term understory plant community responses
to timber-harvesting intensity on non-industrial private forestlands in
Pennsylvania. Forest Ecology and Management 116, 129–139.
89
Gilliam, F.S., 2002. Effects of harvesting on herbaceous layer diversity of a central
Appalachian hardwood forest in West Virginia, USA. Forest Ecology and
Management 155, 33–43.
Grime, J.P., 1977. Evidence for the existence of three primary strategies in plants
and its relevance to ecological and evolutionary theory. The American
Naturalist 111, 1169–1194.
Gurevitch, J., Curtis, P.S., Jones, M.H., 2001. Meta-analysis in ecology. In: Advances in
Ecological Research. Academic Press, pp. 199–247.
Gurevitch, J., Morrow, L.L., Wallace, A., Walsh, J.S., 1992. A metaanalysis of
competition in field experiments. American Naturalist 140, 539–572.
Halpern, C.B., Spies, T.A., 1995. Plant species diversity in natural and managed
forests of the Pacific Northwest. Ecological Applications 5, 913–934.
Halpern, C.B., McKenzie, D., Evans, S.A., Maguire, D.A., 2005. Initial responses of
forest understories to varying levels and patterns of green-tree retention.
Ecological Applications 15, 175–195.
Hedges, L.V., 1981. Distribution theory for glass’s estimator of effect size and related
estimators. Journal of Educational Statistics 6, 107–128.
Harrod, R.J., Povak, N.A., Peterson, D.W., 2007. Comparing the effectiveness of
thinning and prescribed fire for modifying structure in dry coniferous forests.
BW Butler and W. Cook, compilers. The fire environment: innovations,
management, and, policy, 327–346.
Hedges, L.V., Gurevitch, J., Curtis, P.S., 1999. The meta-analysis of response ratios in
experimental ecology. Ecology 80, 1150–1156.
Hedges, L.V., Olkin, I., 1985. Statistical Methods for Meta-Analysis. Academic Press,
New York.
Hooper, D.U., Vitousek, P.M., 1997. The effects of plant composition and diversity on
ecosystem processes. Science 277, 1302–1305.
Jackson, B.C., Pittillo, J.D., Allen, H.L., Wentworth, T.R., Bullock, B.P., Loftis, D.L., 2009.
Species Diversity and Composition in Old Growth and Second Growth Rich
Coves of the Southern Appalachian Mountains. Castanea 74, 27–38.
Jenkins, M.A., Parker, G.R., 1999. Composition and diversity of ground-layer
vegetation in silvicultural openings of southern indiana forests. The American
Midland Naturalist 142, 1–16.
Kern, C.C., Palik, B.J., Strong, T.F., 2006. Ground-layer plant community responses to
even-age and uneven-age silvicultural treatments in Wisconsin northern
hardwood forests. Forest Ecology and Management 230, 162–170.
Kreyling, J., Schmiedinger, A., Macdonald, E., Beierkuhnlein, C., 2008. Slow
understory redevelopment after clearcutting in high mountain forests.
Biodiversity and Conservation 17, 2339–2355.
Lavorel, S., Garnier, E., 2002. Predicting changes in community composition and
ecosystem functioning from plant traits: revisiting the Holy Grail. Functional
Ecology 16, 545–556.
Loya, D.T., Jules, E.S., 2008. Use of species richness estimators improves evaluation
of understory plant response to logging: a study of redwood forests. Plant
Ecology 194, 179–194.
Magurran, A.E., 2004. Measuring Biological Diversity. Blackwell Publishing
Company, Malden, MA.
Meier, A.J., Bratton, S.P., Duffy, D.C., 1995. Possible ecological mechanisms for loss of
vernal-herb diversity in logged eastern deciduous forests. Ecological
Applications 5, 935–946.
Metlen, K.L., Fiedler, C.E., 2006. Restoration treatment effects on the understory of
ponderosa pine/Douglas-fir forests in western Montana, USA. Forest Ecology
and Management 222, 355–369.
Mladenoff, D.J., Pastor, J., 1993. Sustainable forest ecosystems in the northern
hardwood and conifer forest region: concepts and management. In: Aplet, G.H.,
Johnson, N., Olson, J.T., Sample, V.A. (Eds.), Defining Sustainable Forestry. Island
Press, Washington, DC, pp. 145–180.
Moola, F.M., Vasseur, L., 2004. Recovery of late-seral vascular plants in a
chronosequence of post-clearcut forest stands in coastal Nova Scotia, Canada.
Plant Ecology 172, 183–197.
Moola, F.M., Vasseur, L., 2008. The maintenance of understory residual flora with
even-aged forest management: A review of temperate forests in northeastern
North America. Environmental Reviews 16, 141.
Nagaike, T., Kamitani, T., Nakashizuka, T., 1999. The effect of shelterwood logging on
the diversity of plant species in a beech (Fagus crenata) forest in Japan. Forest
Ecology and Management 118, 161–171.
Noss, R.F., 1990. Indicators for monitoring biodiversity: a hierarchical approach.
Conservation Biology 4, 355–364.
Oliver, C.D., Larson, B.C., 1996. Forest Stand Dynamics. John Wiley & Sons, New York.
Outcalt, K. W. 2005. Restoring structure and composition of longleaf pine
ecosystems of the Gulf Coastal Plains. In: Kush, J.S. (Ed.), Proceedings of the
5th Longleaf Alliance Regional Conference, October 12–15, 2004, Hattiesburg,
Mississippi. Longleaf Alliance Report No. 8. Auburn University. Auburn,
Alabama, USA. pp. 97–100.
Paillet, Y., Berges, L., Hjalten, J., Ódor, P., Avon, C., Bernhardt-Romermann, M.,
Bijlsma, R.-J., De Bruyn, L., Fuhr, M., Grandin, U., Kanka, R., Lundin, L., Luque, S.,
Magura, T., Matesanz, S., Meszaros, I., SebastiÀ, M.T., Schmidt, W., Standovar, T.,
Tothmeresz, B., Uotila, A., Valladares, F., Vellak, K., Virtanen, R., 2010.
Biodiversity differences between managed and unmanaged forests: metaanalysis of species richness in Europe. Conservation Biology 24, 101–112.
Parker, G.R., 1993. Limitations on ecosystem management in the central hardwood
region. In: Aplet, G.H., Johnson, N., Olson, J.T., Sample, V.A. (Eds.), Defining
Sustainable Forestry. Island Press, Washington, DC, pp. 202–216.
Pärtel, M., 2006. Data availability for macroecology: how to get more out of regular
ecological papers. Acta Oecologica 30, 97–99.
90
M.C. Duguid, M.S. Ashton / Forest Ecology and Management 303 (2013) 81–90
Peet, R.K., Christensen, N.L., 1988. Changes in species diversity during secondary
forest succession on the North Carolina Piedmont. In: During, H.J., Werner, A.,
Willems, J.H. (Eds.), Diversity and Pattern in Plant Communities. SPB Academic
Publishing, Hague, pp. 233–245.
Phillips, R.J., Waldrop, T.A., 2008. Changes in vegetation structure and composition
in response to fuel reduction treatments in the South Carolina Piedmont. Forest
Ecology and Management 255, 3107–3116.
Quinby, P.A., 1991. Self-replacement in old-growth white pine forests of Temagami,
Ontario. Forest Ecology and Management 41, 95–109.
R Development Core Team, 2012. R: a language and environment for statistical
computing. In: R Foundation for Statistical Computing, Vienna, Austria.
Ramovs, B.V., Roberts, M.R., 2003. Understory vegetation and environment response
to tillage, forest harvesting, and conifer plantation development. Ecological
Applications 13, 1682–1700.
Ritchie, M.W., 2005. Ecological research at the Goosenest Adaptive Management
Area in northeastern California. US Department of Agriculture, Forest Service,
Pacific Southwest Research Station.
Roberts, M.R., Gilliam, F.S., 2003. Response of the herbaceous layer to disturbance in
eastern forests. In: Gilliam, F.S., Roberts, M.R. (Eds.), The Herbaceous Layer in
Forests of Eastern North America. Oxford, New York, pp. 302–320.
Roberts, M.R., Zhu, L., 2002. Early response of the herbaceous layer to harvesting in a
mixed coniferous-deciduous forest in New Brunswick, Canada. Forest Ecology
and Management 155, 17–31.
Ruben, J.A., Bolger, D.T., Peart, D.R., Ayres, M.P., 1999. Understory herb assemblages
25 and 60 years after clearcutting of a northern hardwood forest, USA.
Biological Conservation 90, 203–215.
Scheller, R.M., Mladenoff, D.J., 2002. Understory species patterns and diversity in
old-growth and managed northern hardwood forests. Ecological Applications
12, 1329–1343.
Scherer, G., Zabowski, D., Java, B., Everett, R., 2000. Timber harvesting residue
treatment. Part II. Understory vegetation response. Forest Ecology and
Management 126, 35–50.
Schwilk, D.W., Keeley, J.E., Knapp, E.E., McIver, J., Bailey, J.D., Fettig, C.J., Fiedler, C.E.,
Harrod, R.J., Moghaddas, J.J., Outcalt, K.W., Skinner, C.N., Stephens, S.L., Waldrop,
T.A., Yaussy, D.A., Youngblood, A., 2009. The national fire and fire surrogate
study: effects of fuel reduction methods on forest vegetation structure and
fuels. Ecological Applications 19, 285–304.
Selmants, P.C., Knight, D.H., 2003. Understory plant species composition 30, Äì
50 years after clearcutting in southeastern Wyoming coniferous forests. Forest
Ecology and Management 185, 275–289.
Small, C.J., McCarthy, B.C., 2002. Effects of simulated post-harvest light availability
and soil compaction on deciduous forest herbs. Canadian Journal of Forest
Research 32, 1753–1762.
Smith, D.M., Larson, B.C., Kelty, M.J., Ashton, M.S., 1997. The Practice of Silviculture:
Applied Forest Ecology. John Wiley & Sons, New York.
Stephens, S.L., Moghaddas, J.J., 2005. Silvicultural and reserve impacts on potential
fire behavior and forest conservation: Twenty-five years of experience from
Sierra Nevada mixed conifer forests. Biological Conservation 125, 369–379.
Stewart, G., 2010. Meta-analysis in applied ecology. Biology Letters 6, 78–81.
Takafumi, H., Hiura, T., 2009. Effects of disturbance history and environmental
factors on the diversity and productivity of understory vegetation in a
cool-temperate forest in Japan. Forest Ecology and Management 257, 843–
857.
Thomas, S.C., Halpern, C.B., Falk, D.A., Liguori, D.A., Austin, K.A., 1999. Plant diversity
in managed forests: understory responses to thinning and fertilization.
Ecological Applications 9, 864–879.
Tilman, D., Knops, J., Wedin, D., Reich, P., Ritchie, M., Siemann, E., 1997. The
influence of functional diversity and composition on ecosystem processes.
Science 277, 1300–1302.
Tummers, B., 2006. DataThief III. <http://datathief.org/>.
UNEP, 2010. Global biodiversity outlook 3. In: Diversity, Secretariat of the
Convention on Biological Diversity, Montreal, pp. 94.
Viechtbauer, W., 2010. Conducting meta-analyses in R with the metafor package.
Journal of.
Waldrop, T.A., Yaussy, D.A., Phillips, R.J., Hutchinson, T.A., Brudnak, L., Boerner, R.E.J.,
2008. Fuel reduction treatments affect stand structure of hardwood forests in
Western North Carolina and Southern Ohio, USA. Forest Ecology and
Management 255, 3117–3129.
Wayman, R.B., North, M., 2007. Initial response of a mixed-conifer understory plant
community to burning and thinning restoration treatments. Forest Ecology and
Management 239, 32–44.
Whigham, D.F., 2004. Ecology of woodland herbs in temperate deciduous forests.
Annual Review of Ecology, Evolution, and Systematics 35, 583–621.
Whitlock, M.C., 2011. Data archiving in ecology and evolution: best practices.
Trends in Ecology and Evolution 26, 61–65.
Wood, A., Stedman-Edwards, P., Mang, J. (Eds.), 2000. The Root Causes of
Biodiversity Loss. Earthscan Publications, London.
Wyatt, J.L., Silman, M.R., 2010. Centuries-old logging legacy on spatial and temporal
patterns in understory herb communities. Forest Ecology and Management 260,
116–124.
Yorks, T.E., Dabydeen, S., 1999. Seasonal and successional understory vascular plant
diversity in second-growth hardwood clearcuts of western Maryland, USA.
Forest Ecology and Management 119, 217–230.
Youngblood, A., Metlen, K.L., Coe, K., 2006. Changes in stand structure and
composition after restoration treatments in low elevation dry forests of
northeastern Oregon. Forest Ecology and Management 234, 143–163.