This article was published in an Elsevier journal. The attached copy
is furnished to the author for non-commercial research and
education use, including for instruction at the author’s institution,
sharing with colleagues and providing to institution administration.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Available online at www.sciencedirect.com
Forest Ecology and Management 255 (2008) 1342–1352
www.elsevier.com/locate/foreco
Location relative to a retention patch affects the ECM fungal
community more than patch size in the first season after timber
harvesting on Vancouver Island, British Columbia
Melanie D. Jones a,*, Brendan D. Twieg a,
Daniel M. Durall a, Shannon M. Berch b
b
a
Biology and Physical Geography Unit, UBC Okanagan, 3333 University Way, Kelowna, BC V1V 1V7, Canada
British Columbia Ministry of Forests, Research Branch Laboratory, P.O. Box 9536, Victoria, BC V8W 9C4, Canada
Received 19 July 2007; received in revised form 19 October 2007; accepted 24 October 2007
Abstract
Silviculture systems that include retention of green trees are becoming more common in North America. The goals of green tree retention are to
maintain forest structural diversity, preserve species associated with mature forests and to support faster post-harvest recovery of biodiversity. We
studied the proportion of living roots and ectomycorrhizal (ECM) fungal communities in, and adjacent to, aggregated retention patches of coastal
western hemlock forest on Vancouver Island, 4–6 months after harvest. Our objectives were to determine, for the window of time during which
replanting typically occurs, (i) whether aggregated patches of green trees had retained ECM fungal communities similar to uncut forest and
whether this depended on patch size; (ii) how far the influence of the patch extended into the harvested area, and whether this depended on patch
size. These factors will influence the effectiveness of the aggregated patches as inoculum sources for seedlings planted in adjacent harvested areas.
Soil samples were collected at the center and edge of 16 patches: four replicates each of 5, 10, 20, and 40 m diameter patches, as well as at 10 and
20 m into the harvested area around each patch. A control-forested area was also sampled. The state of the stele was used to designate 25 lateral
roots from each sample as live or dead. One hundred active mycorrhizas per sample were then examined and described morphologically. The
internal transcribed spacer region of the fungal rDNA was amplified and sequenced from representative tips of each morphotype. ECM
communities were indistinguishable between uncut forest and the aggregated retention patches. This was true for patches as small as 5 m in
diameter, with no significant overall effect of patch size on ECM fungal species richness, Shannon Diversity Index, or the proportion of live root
segments. Sampling location, however, significantly affected all these variables, with the influence of the patch disappearing by 10 m into the
harvested area. The only indication of a patch size effect was that ECM species richness at the edges of the 5 m plots was slightly lower (P < 0.1)
than the edges of larger patch sizes. Based on these results, we recommend that patch sizes be at least 10 m in diameter for coastal western hemlock
forests. Since the edge:area ratio of smaller patches is higher, more small patches of at least 10 m diameter would be more effective than a few large
patches in supplying ECM inoculum to adjacent harvested areas during the first year after harvest.
# 2007 Elsevier B.V. All rights reserved.
Keywords: Variable retention harvesting; Ectomycorrhizas; Species diversity; Community composition; Root death; Western hemlock
1. Introduction
Sustained forest productivity and resilience after disturbance
depend on an active soil microbial community. In forest soils,
fungi, including ectomycorrhizal (ECM) fungi, form a
* Corresponding author. Tel.: +1 250 807 9553.
E-mail addresses: [email protected] (M.D. Jones),
[email protected] (B.D. Twieg), [email protected]
(D.M. Durall), [email protected] (S.M. Berch).
0378-1127/$ – see front matter # 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.foreco.2007.10.042
substantial component of the soil microbial community and
play important roles in mineralization of soil organic matter
(Schimel and Bennett, 2004). Clear-cut logging reduces the
fungal:bacteria ratio and changes the species composition of
the ECM fungal community (Kranabetter and Wylie, 1998;
Durall et al., 1999; Hagerman et al., 1999a,b; Siira-Pietikainen
et al., 2001; Twieg et al., 2007). Many ECM fungal species
disappear, while others that are resistant in one way or another
to the disturbance increase in abundance, resulting in a
community shift rather than a reduction in overall mycorrhizal
colonization. In their recent review, Jones et al. (2003) suggest
Author's personal copy
M.D. Jones et al. / Forest Ecology and Management 255 (2008) 1342–1352
two possible explanations for this species shift: (i) that the ECM
fungal communities on seedling roots in clear-cuts are better
adapted to taking up nutrients under the altered conditions or
(ii) that inoculum of some ECM fungi is destroyed during
harvest and so the pool of fungi effectively able to colonize
seedlings is reduced. If the range of fungal species able to
colonize seedlings after harvest is reduced, this may be a reason
for concern because ECM fungi vary greatly in physiology, and
so some major ecosystem functions could be lost (Read and
Perez-Moreno, 2003; Read et al., 2004).
The goal of variable retention harvesting, also known as
green-tree retention, is to preserve species associated with
mature forests, to maintain structural diversity at the landscape
level, and to leave a legacy to support faster post-harvest
recovery of biodiversity (Luoma et al., 2006; Macdonald and
Fenniak, 2007). We expect retention of green trees to fulfill
these goals for ECM fungi, at least during the years
immediately after harvesting (see Peay et al., 2007). For
example, Twieg et al. (2007) have determined that the
mycorrhizal communities in stands of mixed birch/Douglasfir younger than 30-years old differed considerably from those
of 65 years and older. Hence, retention of patches of older forest
would be expected to increase ECM diversity at a landscape
level. Furthermore, variable retention should accelerate the
reintroduction of ECM fungal species into the harvested areas
by increasing inoculum availability to seedlings. This is
because many species of ECM fungi colonize more effectively
from the root systems of other trees than from other forms of
inoculum (Simard et al., 1997; Kranabetter, 1999). Consistent
with this prediction, Outerbridge and Trofymow (2004) found a
reduction in mycorrhizal colonization of Douglas-fir seedlings
with distance away from patches of aggregated retention. In
addition, Luoma et al. (2006) found a 50% reduction in number
of types of Douglas-fir ectomycorrhizas in soils 8–25 m from
individual trees in a dispersed retention system, 1–2 years after
logging. Neither of these studies examined differences amongst
patch sizes or interactions between patch size and distance from
the patch. Hence, we cannot yet recommend the appropriate
size and density of aggregated retention patches needed to
obtain a benefit from variable retention to ECM diversity.
Here we study the effect of variable retention harvesting on
ECM fungal communities in mature second-growth coastal
western hemlock stands on Vancouver Island. The objectives of
this study were to determine, for the window of time during
which replanting typically occurs: (i) whether aggregated
patches of green trees had retained ECM fungal communities
similar to uncut forest and whether this depended on patch size;
(ii) how far the influence of the patch extended into the
harvested area, and whether this depended on patch size. We
examined ECM communities at the center and edge of variable
retention patches ranging in size from 5 to 40 m, as well as up to
20 m into the harvested areas. We expected that the influence of
the patches would extend less than 20 m, based on previous
studies that examined ECM communities along transects from
uncut forests into clear-cuts (Kranabetter and Wylie, 1998;
Hagerman et al., 1999a,b; Cline et al., 2005), from retention
patches into harvested areas (Outerbridge and Trofymow,
1343
2004), or from undisturbed forests into abandoned agricultural
fields (Dickie and Reich, 2005). We further hypothesized that
there would be a minimum patch size below which the species
richness of the ECM fungal community would decrease.
2. Materials and methods
2.1. Site description
The Silviculture Treatments for Ecosystem Management in
the Sayward (STEMS) experiment tests seven treatments:
extended rotation (non-treatment control), extended rotation
with commercial thinning, aggregate and uniform dispersed
retention systems, group selection and modified patch cut
systems, and a 10-ha clear-cut with reserves (de Montigny,
2004). This overall design will be established at three different
areas on northern Vancouver Island during different years. In
this study, we examine the aggregated retention treatment,
control, and clear-cut areas only from the second replicate
(STEMS2). In the Sayward Forest the major stand-initiating
events are believed to have been infrequent fires of moderate
size (20–1000 ha) occurring about every 200 years. Therefore,
the natural forest landscapes would have been dominated by
roughly even-aged stands greater than 20 ha, less than 200years old, with pockets of unburned, old-seral stands greater
than 200 years. Over the past century, logging or fire has
disturbed 63% of the forests in the Sayward Forest (de
Montigny, 2004).
The STEMS2 site is located near Elk Bay, BC, at UTM
323800 E 5576300 N in the CWHxm2 (Coastal Western
Hemlock biogeoclimatic zone, western very dry maritime
subzone, Green and Klinka, 1994). The stand, prior to harvest,
consisted of 81–100 years old (age class 5) western hemlock
(Tsuga heterophylla, 77%), Douglas-fir (Pseudotsuga menziesii, 17%), western redcedar (Thuja plicata, 5%), and grand fir
(Abies grandis, 1%). The site index (average tree height in m at
50 years) for the 25.9 ha study site (called Standards Unit 7 in
the forest management plan) is 37 for western hemlock and 36
for Douglas-fir. The site is located at mid-lower slope position,
aspect N/NE, elevation 37–160 m, and the average slope is
26%. The soil is an Orthic Ferro-Humic Podzol with mormoder
humus form. The forest floor is approximately 5 cm deep and
overlies a well-drained, loamy soil, with approximately 30%
coarse fragment content, and 50 cm rooting depth.
Timber was harvested in mid-March to mid June 2005 using
ground-based mechanized equipment (feller bunchers, feller
processors, and tracked log loaders). Retention patches are
comprised of trees representative of the original stand (western
hemlock, Douglas-fir, western redcedar, and minor amount of
grand fir). Selected patches were spirally pruned to increase
their wind firmness.
2.2. Sample collection
In late October 2005, soil samples were collected at the
center and edge of each patch and at 10 and 20 m along a
transect north from the patch edge into the cutblock. Four
Author's personal copy
1344
M.D. Jones et al. / Forest Ecology and Management 255 (2008) 1342–1352
replicate patches each of 5, 10, 20, and 40 m diameter were
sampled. These patches contained 1–2, 4–10, 9–27, and 32–61
mature, live trees, respectively. In addition, four sampling
points randomly located in an adjacent control forested area
were also sampled. At each sample location, two 4.5 cm
diameter soil cores were collected to 15 cm including forest
floor and upper mineral soil, and bulked into one sample. Thus,
a total of 68 samples were collected and analyzed (four patch
sizes ! four sampling locations ! four replicates = 64 samples, plus four from the control forest). Samples were kept cool
until processed in the lab.
2.3. Morphological analysis of ectomycorrhizas
Ectomycorrhizas were removed from soil samples and
examined morphologically from November 2005 through
May 2006. Roots were gently washed over a 1 mm sieve.
Conifer roots were then moved to a shallow square glass dish
containing deionized water. To determine the proportion of
live versus dead lateral roots, 25 lateral root segments
(approximately 2–7 cm in length) were selected randomly by
identifying a location on a numbered 1-cm grid placed under
the dish. Live lateral roots were defined as having a visible
stele that did not disintegrate when pulled gently with forceps.
Live roots were returned to the dish. Then, all lateral roots
with ectomycorrhizas were cut into 1–1.5 cm pieces. All short
roots were examined on randomly selected pieces until 100
active ectomycorrhizas had been examined. If the sample
contained fewer than 100 active ectomycorrhizas, all active
mycorrhizal root tips were examined. Ectomycorrhizas were
considered to be active if the tips were turgid and the mantle
was intact.
Ectomycorrhizas were classified morphologically using the
detailed approach of Goodman et al. (1996). Root tips were
examined first under a stereoscope; then the mantle was
removed using an ophthalmic scalpel and viewed under 400!
and/or 1000!. Characters such as tip texture, color, length, the
type of mantle pattern, and features of the emanating hyphae
and cystidia were used to place the mycorrhizas into
morphological groups (morphotypes). Photographs were taken
and two representative ECM root tips of each morphotype from
each soil sample were placed into dry, sterile microtubes,
immediately placed at 4 8C, then moved to "80 8C within a few
hours for long-term storage.
2.4. Molecular analysis
DNA was extracted from ectomycorrhizas in CTAB buffer
using a ceramic bead and shaking extractor (Fast Prep FP120,
Holbrook, NY, USA) to pulverize the tissue. DNA was isolated
with chloroform–isoamyl alcohol (24:1), precipitated at "20 8C
in isopropanol, washed twice with 70% ethanol, and eluted in
low-EDTA TE buffer (10 mM Tris–Cl pH 8.0; 0.1 mM EDTA).
Two primer pairs were used to amplify fungal DNA from
ectomycorrhizas. Since Basidiomycetes often dominate ECM
fungal communities, primer pair ITS1f (50 -CTTGGTCATTTAGAGGAAGTAA-30 ) and 4B (50 -CAGGAGACTTGTA-
CACGGTCCAG-30 ) was attempted first. If the first PCR
reaction was unsuccessful, genomic DNA was diluted 1:10
and re-amplified with the same primer pair. If this reaction was
unsuccessful, primer pair NSI1 (50 -GATTGAATGGCTTAGTGAGG-30 ) and NLC2 (50 -GAGCTGCATTCCCAAACAACTC30 ), which amplifies DNA from all fungi, was used, first on
undiluted genomic DNA and subsequently on the 1:10 dilution, if
no product was obtained from the undiluted sample. PCR
products were visualized via electrophoresis in 2% agarose gels
made with ethidium bromide or SYBR-Safe (Invitrogen,
Carlsbad, CA, USA) stain. Samples producing a single PCR
product were cleaned with a Charge Switch PCR Clean-up Kit
(Invitrogen), and cycle sequencing was carried out with a Big
Dye kit (Applied Biosystems, Foster City, CA, USA) using
forward and reverse primers ITS1f and ITS4 (50 TCCTCCGCTTATTGATATGC-30 ) or NSI1 and NLC2. PCR
products that showed multiple bands on gels were not sequenced.
Our goal was to extract fungal DNA from one ectomycorrhiza per
morphotype per sample. To optimize resources, fewer root tips
were processed from morphotypes possessing unique characteristics and well-supported DNA sequences, whereas more
ectomycorrhizas were processed from morphotypes lacking
unique morphological characteristics (e.g. types in the genera
Cortinarius, Russula, and Tomentella).
Forward and reverse sequences were aligned in
Sequencher 4.2 (Gene Codes, Ann Arbor, MI, USA). The
resulting consensus sequences, or single-pass sequences for
cases in which one primer failed to produce a useful
sequence, were then manually corrected and trimmed. Each
sequence was then BLAST-searched through NCBI-linked
and UNITE databases. For NCBI searches, filters were set to
attempt to remove matches to fungal sequences from
mycorrhizal root tips or soil samples; this allowed matching
of our root tip samples to sequences obtained mostly from
fungal sporocarps. A functional species level match criterion
was set at 98% sequence similarity for double-pass and 97%
for single-pass sequences due to base call error in single-pass
sequences (Izzo et al., 2005). If no species level match was
made between a sample sequence and a database sequence,
then the 10 highest-scoring BLAST matches were checked
and taxonomic placement of the sample was made according
to the lowest possible taxonomic consensus (i.e. genus,
family, or order) of those matches. Sample sequences were
grouped to the genus or family level according to BLAST
results, and separate alignments were performed on each
group in ClustalX (Thompson et al., 1997). Pair-wise
sequence similarities were then calculated within each group
(ignoring gaps and unidentified bases), and samples were
grouped at the species level according to the same criteria
used in matching to database entries via BLAST searches.
Alignments were then visually reviewed to assure quality
and detect potentially spurious species matches due to
numerous gap insertions, large numbers of unidentified
bases, and/or short sequences that spanned the entire nonvariable 5.8 s region. Sequences that could not be matched to
one unique sequence type without ambiguity were not used
for analyses.
Author's personal copy
M.D. Jones et al. / Forest Ecology and Management 255 (2008) 1342–1352
1345
Table 1
List of taxa identified by DNA sequences analysis and/or morphotyping, pertinent information from BLAST searches, and average relative abundances of these taxa
by sampling location
Functional ‘‘species’’ name
Ascomycetes 1
Atheliaceae 1
Atheliaceae 2
Atheliaceae 3
Atheliaceae 4
Amphinema genus total
Basidiomycetes 1
Boletus zelleri
Cenoccoccum geophilum
Chamonixia cf. ambigua
Clavulina cinerea
Clavulina 1
Cortinarius/Dermocybe 1
Cortinarius acutus
Cortinarius angelesianus
Cortinarius aureovelatus
Cortinarius cf. iliopodius
Cortinarius rubellus
Cortinarius 1
Cortinarius 2
Cortinarius 3
Cortinarius 4
Cortinarius 5
Cortinarius 6
Cortinarius 7
Cortinarius genus total
Craterellus tubaeformis
Elaphomyces granulatus
‘‘Gelwalls’’
Hygrophorus 1
Inocybe cf. calamistrata
Inocybe 1
Inocybe 3
Inocybe 6
Inocybe genus total
ITE-6-like
Laccaria bicolor
Lactarius 1
Lactarius 2
Lactarius scrobiculatus
Lactarius genus total
MRA
Piloderma 1
Piloderma 2
Piloderma genus total
Rhizopogon genus total
Russula emetica
Russula cf. favrei
Russula nigricans
Russula vinosa
Russula xerampelina
Russula 1
Russula 2
Russula 3
Russula genus total
Suillus genus total
Thelephorales 1
Thelephorales 2
Thelephorales 3
Closest database
match(es)c
Sphaerosporella brunnea
Piloderma fallax
uncultured ECM tip
P. fallax
Amphinema byssoides
Morphotype only
Clavulinaceae sp.
Boletus zelleri
Morphotype only
Chamonixia cf. ambigua*
Clavulina cinerea
Clavulina cristata
Cortinarius amoenus
Cortinarius acutus
Cortinarius angelesianus
Cortinarius aureovelatus
Cortinarius cf. iliopodius
Cortinarius rubellus
Cortinarius ochrophyllus
Cortinarius obtusus
Cortinarius cf. saniosus
Cortinarius cagei
Cortinarius obtusus
Cortinarius sp.
Cortinarius alboviolaceus
Morphotypes included
Craterellus tubaeformis
Elaphomyces granulatus*
Morphotype only
Hygrophorus camarophyllus
Inocybe cf. calamistrata*
Inocybe cf. glabripes
Inocybe nitidiuscula
Inocybe aurea
Morphotypes included
Morphotype only
Laccaria bicolor
Lactarius repraesentaneus
Lactarius lucentulus
Lactarius mitissimus
Morphotype only
Morphotypes included
Morphotype only
Piloderma sp. B22
Piloderma olivaceum
Includes P. fallax morphotype
Morphotypes only
Russula emetica
Russula cf. favrei
Russula nigricans
Russula vinosa
Russula xerampelina
Russula crassotunicata
Russula xerampelina
Russula raoultii
Morphotypes included
Morphotypes only
Pseudotomentella tristis
Pseudotomentella tristis
Pseudotomentella tristis
Accession
numbera
UDB000994
DQ365674
AY394895
UDB001614
AY219839
n/a
AJ534710
AY750158
n/a
AF335456
AY292292
AF389160
UDB001547
UDB000671
UDB001768
AJ889948
AY669595
UDB001127
AJ238035
DQ102683
AY669676
UDB000127
UDB001543
DQ097877
n/a
AF385632
n/a
UDB000561
AJ889952
AB244791
UDB000612
n/a
n/a
DQ179121
UDB000367
DQ384582
AF157412
n/a
n/a
n/a
AJ534903
UDB locked
n/a
n/a
AY061673
DQ384579
AM113961
AY061724
DQ367916
DQ384530
DQ367916
AF518621
n/a
n/a
UDB000279
UDB000279
AJ8899968
Base pairs
alignedb
Similarity
(%)
157
251
778
218
579
n/a
242
568
n/a
#600
634
545
472
565
492
624
595
697
374
627
601
666
622
576
545
n/a
716
#700
n/a
428
#450
295
271
249
n/a
n/a
642
715
642
638
n/a
n/a
n/a
687
99
96
99
93
96
n/a
95
99
n/a
99
99
96
90
99
99
99
98
99
97
95
96
97
96
96
95
n/a
100
98
n/a
93
99
92
92
91
n/a
n/a
98
95
99
98
n/a
n/a
n/a
100
n/a
n/a
545
643
559
531
704
350
658
686
n/a
n/a
502
487
573
n/a
n/a
99
99
99
99
99
97
93
96
n/a
n/a
96
88
90
Relative abundance (%) by sampling
location
Center
Edge
10 m
20 m
Forest
0
0.8
0.3
0.4
0
0
0
0
24.7
1.7
0.5
0.7
0.4
0
<0.1
0
0.2
0
0
4.0
0
0.1
4.6
1.0
0.3
13.9
0.1
0
0.7
0.5
0
1.6
1.2
0.5
6.8
0
0
0
<0.1
0.1
0
0
0.1
1.0
0
0
0.3
21.7
0
0
0
0
0
0
0.4
0
0
1.7
0
0
0
0
0
0
6.4
0
0
3.6
0
0
0.4
0
0
4.6
1.3
0.4
0.5
1.1
0
0
0
0
0
0
0
0
34.8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.5
0
0
0
0
2.1
0
0
0
4.0
0
0
0
0
0
0
0
0
0
0
0
0
44.2
0
0
0
0
0
0
0
0
<0.1
0
0
0
0
0
0
0
1.0
0
0.1
0
0
0.3
1.0
0
0
4.1
0
0
0
0
0
0
0
0
0
0.8
1.3
0
21.7
0
17.7
0
0
4.3
0
0
0
0
0
0
1.3
0
0
0
0
5.8
0
0
4.8
0
0
0
0
0
0
0
0
0
0
0.9
0.9
6.0
1.4
0
6.0
1.4
1.6
0
0
4.2
1.8
0.5
0.3
1.7
23.7
0
0.3
1.0
0
3.7
5.8
3.1
0
2.0
3.1
1.3
0.2
0
0
0
3.0
0
0
0.1
15.2
4.4
0
0.3
0.2
0
1.0
10.9
0
0.2
11.0
0
0
0
0
0
0.2
0
0
0
12.4
6.1
0
0
0
0
0
11.7
0
0
11.7
0.5
0
1.4
3.6
0
1.9
0
0
0
18.7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
20.4
4.3
0
0
0
Author's personal copy
1346
M.D. Jones et al. / Forest Ecology and Management 255 (2008) 1342–1352
Table 1 (Continued )
Functional ‘‘species’’ name
Thelephorales 4
Tomentella 1
Tomentella 2
Tomentella genus total
Tricholoma atroviolaceum
Tricholoma sejunctum
Tylospora 1
Tylospora 2
Tylospora genus total
‘‘Yellow Unknown’’
a
b
c
Closest database
match(es)c
Pseudotomentella vepallidospora
Tomentella badia
Tomentella ramosissima
Morphotypes included
Tricholoma atroviolaceum
Tricholoma sejunctum
Tylospora asterophora
Tylospora fibrillose
Morphotypes included
Morphotype only
Accession
numbera
AF374773
UDB000961
U83480
n/a
AY750166
AF377192
AF052557
AF052563
n/a
n/a
Base pairs
alignedb
500
465
494
n/a
480
658
639
618
n/a
n/a
Similarity
(%)
93
96
94
n/a
99
99
96
95
n/a
n/a
Relative abundance (%) by sampling
location
Center
Edge
10 m
20 m
0
0.7
0
1.5
0
0
0.4
0.9
1.3
9.0
1.0
0
2.1
6.0
0
0.3
0
1.3
6.3
8.5
0
0
0
1.9
0.3
1.4
0
0.3
0.3
11.7
0
0
0.4
1.6
0
0
0
0
1.6
8.9
Forest
0
0
0
0
0
0
0
0
0
2.8
Accession numbers are for closest database matches.
‘‘Base Pairs Aligned’’ refers only to the longest aligned segment from matches in which unaligned gaps occurred.
Sequences obtained from sporocarps collected and identified from the study site are identified by an asterisk (*).
2.5. Statistical analyses
For analyses of ECM fungal species richness and diversity,
all morphotypes were included that either were identified by
DNA sequences or presented morphological characteristics that
were sufficient to classify them as unique within a sample.
Some morphotypes, for which no DNA sequences were
obtained, were placed into genus or family based on
morphology (for analyses at the genus level and richness/
diversity as described below). These groupings included
Amphinema, Cortinarius, Inocybe, Lactarius, Piloderma,
Russula, Rhizopogon, Suillus, Tomentella, and Tylospora.
However, within these genera, species level identifications
were not made based on morphology alone, except for the rare
occurrence (seven in total) of two obviously different
morphotypes in one genus within a sample (e.g. one Russula
sp. with a regular synenchyma outer mantle and another
Russula sp. with a net prosenchyma outer mantle). These were
included as distinct species for richness and diversity
calculations of the corresponding sample, but were removed
for analyses involving similarity measures of fungal species
composition.
Because many samples contained fewer than 100 active
ECM root tips, and because some tips had to be omitted because
they could not be distinguished through morphological or
molecular methods, ECM fungal species richness was adjusted
by rarefaction (Simberloff, 1972), using a web-based Rarefaction Calculator (J. Brzustowski, http://www2.biology.ualberta.ca/jbrzusto/rarefact.php; accessed March 2007). The fungal
richness of each sample was adjusted to the lowest average
number of ECM tips per sample for any combination of patch
size and sampling location (45 root tips). The Shannon
Diversity Index for each sample, plus species-sample unit
curves and first-order jackknife estimates of species richness
for each sampling location were calculated in PC-ORD Version
4 (MJM Software Design, Gleneden Beach, OR, USA).
Relative abundances of fungal taxa were calculated by dividing
the number of root tips colonized by each taxon in each sample
by the total number of active ECM tips examined in that sample
(up to 100). Taxon frequencies were calculated by totaling the
number of samples (out of four) in which each taxon was found
for each unique combination of patch size and sampling
location.
Analysis of variance (ANOVA) was conducted in SAS
Version 9.1 (SAS Institute, Carey, NC, USA). To test the effects
of patch size and sampling location (objective (ii) and the
second part of objective (i)), the study was analyzed as a split–
plot (two-way) ANOVA, with patch size as the main treatment
effect and sampling location (patch center, patch edge, 10 m
into cut, and 20 m into cut) as the subplot effect, in a completely
randomized design. To test whether or not the patches of
retained trees were different than the uncut control forest (first
part of objective (i), separate one-way ANOVAs were done
using the uncut forest, patch center, and patch edge as
treatments, inclusive of all patch sizes (also a completely
randomized design). We excluded the cut locations from this
analysis because the objective was to determine whether or not
the patches themselves had retained diversity similar to uncut
forest, and the effects of patch size and sampling location were
better analyzed separately by split–plot design, independent of
uncut forest observations. Several variables were tested in
separate ANOVAs using the GLM procedure: ECM fungal
species richness, Shannon Diversity Index, the proportion of
root segments that were dead, and the relative abundances of
fungal taxa having an average relative abundance of at least
10% in one or more patch size/sampling location combination
or at least 5% overall relative abundance. Bonferroni multiple
comparisons were performed where significant effects
(a = 0.05 unless otherwise stated) were found. Relative
abundance data for two taxa violated assumptions of normality
and variance homogeneity, even after arcsine-square-root
transformation, so Kruskal–Wallis tests were run for these
taxa with patch size and sampling location effects addressed in
separate tests.
To assess effects of patch size and sampling location on
ECM fungal community composition, nonmetric multidimensional scaling (NMS) ordinations, and multiple response
permutation procedures (MRPP) were run in PC-ORD. These
were run at two levels of taxonomic grouping. The first was the
species level, which included only those mycorrhizas formed
Author's personal copy
M.D. Jones et al. / Forest Ecology and Management 255 (2008) 1342–1352
by fungi identified to species; morphotypes that could only be
identified to higher taxonomic levels were omitted from this
analysis. The second was the genus level, which included all
sequence types, and morphotypes grouped into the lowest
unique taxonomic unit possible above the species level. Taxon
frequencies for each unique combination of patch size and
sampling location were used, with the Sorensen distance
measure, for both ordinations. The uncut forest was also
included. Monte Carlo tests were performed with 40 NMS runs
on the real data and 500 randomizations to determine the
significance of the ordination structure. PC-ORD chose the best
dimensionality out of six dimensions according to these
significance tests and the number of axes beyond which
ordination stress reduction was minimal. Separate MRPP
procedures were done with patch size and sampling location as
grouping variables.
1347
rarefaction in analysis compensated for the difference in DNA
success among locations when evaluating the effects of patch
size and sampling location on species richness.
Seven morphotypes, for which no DNA sequences were
obtained, were considered unique enough in their morphology
to designate them as unique species within the data set. These
included morphotypes matching the descriptions of mycorrhizas formed by Cenococcum geophilum, ITE-3 (called
‘‘MRA’’ here), and ITE-6 (Ingleby et al., 1990), as well as
those formed by Lactarius scrobiculatus (Kernaghan and
Berch, 1996) and Piloderma fallax (Goodman and Trofymow,
1996). Two other morphotypes were unique, and we could not
find comparable published descriptions (see Supplementary
Appendix A for descriptions).
3.2. Comparison between uncut forest and aggregated
retention patches
3. Results
3.1. Mycorrhizal fungal identification
Including the uncut forest plots, a total of 138 sequences of
acceptable length and quality were obtained, but forty-five of
these were identified as non-target sequences (sequences from
fungi inconsistent with sample morphotype at the genus level or
higher taxonomic level) coming from other fungi co-inhabiting
root tip samples. Most non-target sequences were from
saprotrophic taxa. However, sequence type ‘‘Inocybe 1’’,
which was obtained from two root tip samples showing ECM
morphology consistent with the genus Inocybe (Table 1), was
also amplified from several samples of morphological types
inconsistent with Inocybe. These samples were placed
taxonomically by morphotype. Another 30 samples produced
sequences with double peaks, indicating that a second fungus,
which was not visible during electrophoresis, was co-amplified
in the PCR. Fifty-three unique sequence types (hereafter
referred to as ‘‘species’’) were found (given in Table 1 along
with relative abundance). There was only one instance in which
DNA sequences identified multiple species of the same genus
within one soil sample, and there were four instances in which
the same sequence type (i.e. species) was obtained from two
separate ECM systems within the same soil sample when
morphology of those systems was ambiguous. One representative of each sequence type has been submitted to Genbank.
The percentage of samples that provided a single product
from PCR amplification was 86% for the center sampling
location, 81% for the edge, 70% 10 m into the harvested area,
and 63% 20 m into the harvest area. Due to a high proportion of
samples either producing sequences of non-target or saprotrophic taxa, or having two peaks in the sequence output due to
the undetected presence of DNA of a second fungus, the
percentage of samples that provided a useful DNA sequence
from the target ECM fungus out of the total samples on which
PCR was attempted was 54% for the patch center, 38% for the
patch edge, 25% for 10 m into the cut, and 33% for 20 m into
the cut. There was only a 5% range in this DNA success among
patch sizes, and the inclusion of morphotype data and the use of
The aggregated retention patches appeared to retain forestlike characteristics with respect to the ECM fungal community
and root vitality. Soil samples collected at the patch centers and
edges had similar mean ECM fungal species richness, Shannon
Fig. 1. Mean (n = 16 except n = 4 for uncut forest) ECM fungal species
richness (a) and proportion of root segments that were dead (b) by sampling
location. Error bars represent one standard error of the mean. Means with the
same letter are not significantly different. Uncut forest (diagonal lines through
bars) were not part of analyses for mean comparisons.
Author's personal copy
1348
M.D. Jones et al. / Forest Ecology and Management 255 (2008) 1342–1352
Fig. 2. Species-sample unit curves by sampling location: 20 m into cut (empty
triangles), 10 m into cut (filled triangles), patch edge (empty circles), patch
center (filled circles) and control forest (empty squares). Numbers next to
curves are first-order jackknife estimates of species richness; no estimate is
given for uncut forest because the number of samples is central to the
calculation, and the estimate is therefore not comparable to the other sample
locations.
Diversity Index, and proportion of dead root segments as
samples from the uncut forest (Fig. 1a; Table 2; Supplementary
Fig. 1a) This suggests that, at the sample scale, the patches had
not lost a significant number of ECM fungi since harvest. As far
as they can be compared, species accumulation curves were
similar for communities in patches and in the uncut forest
(Fig. 2). However, only four soil samples were collected from
the uncut forest, compared to 16 samples from the center and
edges of the patches (including all patch sizes), so we can
compare the rate of species accumulation for only the lower end
of the curve. Approximately 50% of the lateral roots sampled
were dead in all forested samples.
3.3. Effects of patch size
Patch size had no overall effect on mean rarefaction-adjusted
species richness or on Shannon Diversity Index of the ECM
fungal community; nor did the proportion of dead root
segments vary with patch size when data was combined across
Fig. 3. Mean (n = 4) ECM fungal species richness by combination of patch size
and sampling location; 5 m patch (filled circles), 10 m patch (empty circles),
20 m patch (filled triangles), and 40 m (empty triangles). Bars represent one
standard error of the mean.
sampling locations (Table 2). However, average diversity and
richness were lower at the edges of the 5 m patches than other
patch sizes, explaining why the interaction between patch size
and sampling location was marginally significant at a = 0.1
(Fig. 3; Table 2 ; Supplementary Fig. 1b). There was no
grouping according to patch size in MRPP or ordination
analysis either at the species (P = 0.32; Fig. 4a) or genus
(P = 0.67; Fig. 4b) level. The only major taxon that showed
differences among patch sizes was the genus Piloderma, which
was of similar relative abundance (% of ECM tips) in 5, 10, and
20 m patch sizes but far lower relative abundance in the 40 m
patch size (Table 3).
3.4. Effects of sampling location
Sampling location significantly affected all three variables,
with a major disjunction between samples collected within the
patches and those collected in the harvested areas. Mean
rarefaction-adjusted species richness was at least 46% higher at
both the patch center and edge locations than 10 and 20 m into
the harvested areas (Fig. 1a). Differences in the Shannon
Table 2
Analysis of variance results comparing species richness or diversity of ECM fungi and the proportion of dead roots based on a one-way analysis with uncut forest,
patch center, and patch edge as treatments (all patch sizes included), and a split–plot analysis with patch size as the main treatment effect and sampling location (patch
center, patch edge, 10 and 20 m into cut) as the subplot effect
Source of variation
d.f.
One-way (w/uncut forest)
Sampling location
2
Error
33
Split–plot design
Patch size
Sampling location
Patch size !
sampling location
Error
Species richness (rarefaction-adjusted)
Shannon diversity index
MS
MS
F
P
F
Proportion of root segments dead
P
MS
F
P
2.054
3.652
0.56
0.5753
0.1250
0.1410
0.89
0.4217
0.0155
0.0169
0.92
0.4088
3
3
9
0.953
33.03
4.895
0.37
12.23
1.81
0.7765
<0.0001
0.0998
0.0305
1.800
0.2808
0.29
11.67
1.82
0.9946
<0.0001
0.0983
0.0195
0.1093
0.0112
0.62
9.66
0.99
0.6169
<0.0001
0.4675
36
2.701
0.1543
0.0113
Author's personal copy
M.D. Jones et al. / Forest Ecology and Management 255 (2008) 1342–1352
Fig. 4. NMS ordinations of combinations of patch size (diameter in meters
indicated by the number directly to the left of each point) and sampling
locations (empty triangles = 20 m into cut; filled triangles = 10 m into cut;
empty circles = patch edge; filled circles = patch center; square = forest control) according to frequency of fungal species (a) and genera (b). R2 is the
proportion of variance in the Sorensen distance matrix among points that is
explained by the axes.
1349
Diversity Index among sampling locations were similar to those
seen in species richness (Supplementary Fig. 1a). The mean
proportion of root segments that were dead at 10 and 20 m into
the cut was 30–37% higher than at the center and edge of the
retention patches (Fig. 1b). Species-sample unit curves show
that the sum of species found for all replicates was far higher at
the patch centers and edges than the locations in the cut (Fig. 2).
It also appears, although none of the sampling locations were
sampled thoroughly enough to account for all species present,
that the number of species accumulated much more rapidly
with added sampling effort at the patch center and edge than the
cut locations.
The species composition of the ECM fungal community also
differed among sampling locations ( p = 0.043; MRPP), but the
effect size was fairly low (A = 0.022). The ordination at the
species level provided three significant axes ( p = 0.0259,
0.0359, and 0.008 for axes 1–3, respectively). The two axes
explaining the greatest proportion of variation in the distance
matrix from each ordination are presented in Fig. 4. Six out of
the eight combinations of patch size and sampling locations at
10 and 20 m into the harvested area grouped together at the
center of axis 3 and the right side of axis 1 while the patch
center and edge locations and uncut forest were placed in
smaller groups near the extreme top and bottom of axis 3 and
the left side of axis 1 (Fig. 4a). At the genus level, MRPP also
showed an effect of sampling location on ECM community
structure ( p = 0.022) and a larger effect size (A = 0.8). The
ordination at the genus level also provided three axes
( p = 0.004, 0.002, and 0.004 for axes 1–3, respectively). A
high proportion of the variation in the distance matrix was
explained by axis 2 (R2 = 0.80; Fig. 4b). Although there were
not strong groupings by sampling location, the uncut forest,
patch center, and patch edge locations were placed generally at
the top half of axis 2 while the locations in the harvested area
were placed in the bottom half.
Two ECM fungal taxa differed in relative abundance or
frequency between the retained patches and the harvested areas
Table 3
Mean frequencies (number of occurrences in the four samples per combination of patch size and sampling location; n = 4) and relative abundances (n = 16; $standard
error of the mean) of the most abundant fungal taxa, by sampling location and patch size
Location
Frequency/relative abundance (%)
Cenococcum geophilum
Cortinarius
Inocybe
Piloderma
Russula
‘‘Yellow unkown’’
Center
Edge
10 m
20 m
4/24.7 $ 5.0 a
4/21.7 $ 4.0 a
3.5/34.8 $ 6.9 ab
3.5/44.2 $ 7.7 b
2/13.9 $ 5.4 b
1/6.4 $ 3.1 ab
0.25/0.54 $ 0.54 a
0.25/1.0 $ 0.69 a
1.5/6.8 $ 2.8
1.5/4.6 $ 2.3
0.5/4.0 $ 1.1
1/4.1 $ 2.5
1.25/6.0 $ 4.0
1/3.1 $ 2.0
1/11.0 $ 6.6
1.5/11.7 $ 5.9
3.25/23.7 $ 6.4
2.75/15.2 $ 6.2
2.25/12.4 $ 6.1
1.75/18.7 $ 5.4
1.5/9.0 $ 4.0
1.25/8.5 $ 4.1
1.75/11.7 $ 6.3
2.25/8.9 $ 3.1
P-value
0.0092
0.0063
0.7467
0.5308*
0.5113
0.7758*
3.75/27.3 $ 5.4
3.75/26.2 $ 5.6
3.75/33.9 $ 6.7
3.75/38.0 $ 7.6
0.75/4.0 $ 3.5
1.25/9.8 $ 4.7
1.25/2.9 $ 1.6
0.75/5.1 $ 2.9
2/10.5 $ 4.5
1/4.0 $ 2.5
0.75/2.6 $ 1.5
0.75/3.3 $ 2.0
1.25/11.2 $ 6.7
2.5/10.0 $ 4.2
1.5/10.2 $ 6.0
0.5/0.44 $ 0.38
2.75/23.5 $ 7.3
3.5/16.7 $ 5.3
2.75/14.5 $ 5.8
2.5/15.3 $ 5.6
1.25/6.8 $ 4.0
1.75/7.3 $ 2.8
2.25/13.1 $ 4.1
1.5/11.0 $ 6.3
0.5458
0.5821
0.4675
0.0360*
0.8589
0.4515*
Patch size
5m
10 m
20 m
40 m
P-value
P-values are for relative abundance, from two-way ANOVA tests except the two with an asterisk (*), which are from a Kruskal–Wallis analysis. Means in columns
with the same letter are not significantly different at / = 0.1.
Author's personal copy
1350
M.D. Jones et al. / Forest Ecology and Management 255 (2008) 1342–1352
(Table 3). Cenococcum geophilum was present at higher
relative abundance within the harvested area than it was at the
patch centers and edges. Its frequency (number of samples in
which it was detected), however, was similar among sampling
locations and among patch sizes. The genus Cortinarius
decreased in relative abundance from the patch center outward
into the harvested area, and its frequency showed a similar
pattern. The other examined taxa showed no differences among
sampling locations.
4. Discussion
In this study, we found no difference in the ECM fungal
community, expressed at the scale of a soil sample, between
patches of green trees and uncut forest regardless of patch size,
but we found marked differences between patches and
harvested areas. Although this indicates that the aggregated
retention system has structural and biological diversity at the
scale of the cutblock, it is important to remember that samples
were collected only 4–6 months after harvest. Island
biogeography theory predicts that the number of ECM fungal
species will decrease in the patches over time as a ‘retraction
flora’ develops (MacArthur and Wilson, 1967). Peay et al.
(2007) recently found that this can occur for ECM communities
as soon as 11 years after wildfire established patches of Pinus
muricata of different sizes. Clearly it was too early for the
patches we studied to exhibit a decrease in richness, but it will
be important to monitor them over time. Ectomycorrhizal
fungal species richness will likely decrease more rapidly in
smaller patches than larger patches according to wellestablished species–area relationships (Diamond, 1975; Simberloff and Abele, 1976). Although it is still a matter of debate
about how closely these rules will apply to micro-organisms
(Green and Bohannan, 2006), Peay et al. (2007) found
significantly higher ECM fungal species richness in larger
pine islands than smaller islands.
The conclusions presented here are based on mean richness
per sample; we cannot compare cumulative species numbers
between uncut forest and the patches because we did not sample
as intensely in the uncut forest as in the retention patches.
However, we would expect lower overall ECM fungal richness
in forest portions of aggregated retention systems than in uncut
forests. Genets of ECM mycelia are known to occur in patches
ranging from less than 1 m2 (Gherbi et al., 1999) to 300 m2 or
more (Bonello et al., 1998) and only some of these are retained
within randomly located green-tree patches. Because of this
high spatial variability, if large patches and small patches were
sampled at the same intensity per unit area, there would
undoubtedly be more ECM fungal species detected overall in
large patches.
Our results clearly show that, on a per sample basis, location
relative to a green-tree patch is much more important than patch
size in determining the proportion of live roots and
ectomycorrhiza diversity in root systems from the harvested
stand. The results regarding the effect of location relative to
living trees are consistent with studies that have examined
ectomycorrhiza diversity on root systems remaining in forest
soils after harvesting (Hagerman et al., 1999b; Outerbridge and
Trofymow, 2004; Cline et al., 2005; Luoma et al., 2006). For
example, Hagerman et al. (1999b) found a reduction from 10
morphotypes per plot in the first summer after winter clear-cut
logging of a sub-alpine spruce-fir forest to two morphotypes per
plot in the third summer. This reduction occurred in plots 16 m
or more into the harvested area. Similar results have been found
recently in dispersed retention systems. Luoma et al. (2006)
found a 50% reduction in number of types of Douglas-fir
ectomycorrhizas 8–25 m from retention trees 1–2 years after
logging. Loss of ECM diversity may have occurred faster in the
STEM sites than in the subalpine ESSF sites studied by
Hagerman et al. (1999a), because warmer temperatures at the
lower elevation encouraged higher metabolic rates in roots; the
CWH zone has 4–6 months per year when the mean
temperature is above 10 8C while the ESSF has two or less,
and their mean annual temperatures are 8 and 08C, respectively
(Meidinger and Pojar, 1991). Given that soil temperature is
strongly linked to root respiration and decomposition
(Biondini, 2001), this could have resulted in a faster
consumption of stored carbohydrates and more rapid death
of roots and fungi.
The numerous comparisons at the patch size ! sampling
location interaction level resulted in a very low adjusted alpha,
so we could not find statistical differences in species diversity
among different patch sizes at the edge location. However, there
appears to be a difference in the relationship between center and
edge locations in 5 m patches as compared to larger patches.
The relationship between these two locations in 10–40 m
patches is consistent with the spatial between-patch model
under intermediate disturbance (Roxburgh et al., 2004), where
species of lower competitive ability are able to coexist at patch
edges with species well adapted to competing in undisturbed
forest. We would expect that the center of a 5 m patch may
indeed be more similar to the edge of a larger patch, in the sense
that the number of living tree roots from different individuals
that coexist in a particular location would be similarly reduced.
It is possible that the density of live roots remaining at the edge
of 5 m patches is lower than that of larger patch sizes, resulting
in lower fungal diversity.
It is surprising that Piloderma was of lower relative
abundance in the largest patch size, since Smith et al. (2000)
and Twieg et al. (2007) both found P. fallax to be far more
frequent in mature stands than young ones. However, the
youngest stands in those studies were 30–35 and 4–6 years old,
respectively, so this comparison is not analogous to our study.
Furthermore, Smith et al. (2000) also found P. fallax occurrence
to be well predicted by the occurrence and cover of highly
decayed coarse woody debris (CWD), which likely results in
Piloderma being quite patchily distributed in accordance with
its preferred substrate. Since we did not examine CWD in our
study, the patch size effect that we found for this fungus may be
an artifact of different amounts of this substrate being present at
our sampling locations.
The proportion of live lateral roots dropped significantly
from the center and edge of the patches to 10 or 20 m into the
harvested area. Parsons et al. (1994) found that the number of
Author's personal copy
M.D. Jones et al. / Forest Ecology and Management 255 (2008) 1342–1352
live fine roots of lodgepole pine dropped substantially further
than 5 m into clear-cuts. Therefore, our samples collected at 10
and 20 m into the harvested area likely contained primarily
dying roots from the harvested stand. However, the proportion
of live roots attached to planted seedlings can also be lower with
increased distance from retained trees (Cline et al., 2007), so
roots extending from patch trees into the harvested area may
also have lower survival rates than those in the patches. The
roots of harvested trees would also have extended into the
patch, thereby resulting in the intermediate proportion of live
roots found at the edges of the patch.
In the study by Luoma et al. (2006), the number of ECM
types was correlated with the reduction in numbers of fine roots
with distance from the living tree. In our study we used
rarefaction to correct for differences in the number of active
ectomycorrhizas per sample, yet we still observed a reduction
in richness in the harvested areas. This suggests that some types
of mycorrhizas persist longer than others as root systems die
following harvest. In this regard, our results agree remarkably
well with those of Hagerman et al. (1999a). In the second
growing season after logging, like us, these authors found that
Cenococcum mycorrhizas formed a higher proportion of the
mycorrhizal community in harvested areas and Cortinarius
mycorrhizas formed a lower proportion in harvested areas than
they did within the forest. In both cases, however, it is possible
that the Cenococcum results are artifacts. Cenococcum
geophilum forms thick, tough, black mantles. As a result, it
is possible that the root tips remain turgid even when the root
and the fungus are dead. The high concentrations of phenolics
in the Cenococcum tissue may impede decomposition of these
tips compared to other root tips. We have evidence to support
this hypothesis: we had a substantially lower success rate in
amplifying DNA of ECM fungi when roots were sampled 10 m
or more from the patch edge, even if those root tips were
characterized as active during the morphotyping process.
Instead, we frequently amplified DNA of saprotrophic fungi,
suggesting that a higher proportion of these root tips were
moribund, even though they fit the morphological criteria of
being ‘active’. It is therefore possible that we overestimated
richness of ECM fungi in the cut locations, as we included
morphotype information when DNA amplification or sequencing was unsuccessful.
5. Conclusions and management implications
For ECM fungal communities, aggregated retention harvesting will retain patches of forest-like communities, for at least 6
months after harvest. This is significant because seedlings are
typically replanted in the first growing season after harvest.
Because we saw a slight effect of patch size on the fungal
community at the edge of the patches, we recommend that
patch size for aggregated retention patches be 10 m in diameter
or greater, when being used as a potential inoculums source for
planted seedlings. Our results suggest that the ‘shadow’ cast by
these patches is quite small, less than 10 m into the harvested
area. Taken together with the findings that colonization of
seedlings decreases with distance from green tree patches
1351
(Outerbridge and Trofymow, 2004) and forest edges (Hagerman et al., 1999b), we conclude that including a large amount
of forest edge in the cutblock will maximize inoculum
availability to seedlings planted in the cutblock. This would
be achieved by leaving a larger number of relatively small
patches than fewer larger patches. This conclusion is valid only
if seedlings are planted relatively soon after harvest because the
ECM fungal community in smaller patches is likely to decrease
in richness more rapidly with time than communities in larger
patches (Peay et al., 2007). Continued study of this system is
required in order to determine the long-term consequences of
aggregated patch size on ECM richness, especially at the
landscape scale. While diversity of ECM fungi on root tips may
be conserved in smaller patches shortly following harvest, we
acknowledge that those smaller patches of trees may be less
likely to survive in the long run due to windthrow (Jönsson
et al., 2007).
Acknowledgements
Funding for this study was provided by the British Columbia
Forest Science Program of Forestry Innovation Investment Ltd.
to Sue Grayston (PI) for the project entitled ‘‘Green Tree
Retention: A tool to maintain ecosystem health and function’’.
We acknowledge the major contribution of Louise deMontigny
of the British Columbia Ministry of Forest and Range as Project
Leader of the STEMS study. We are grateful to Valerie Ward for
morphotyping and DNA analysis, Jeff Sherstobitoff for field
assistance, and Robin Modesto, coordinator of the STEMS
study with International Forest Products, for providing access
to the study site and providing detailed site information.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at doi:10.1016/j.foreco.2007.10.042.
References
Bonello, P., Bruns, T.D., Gardes, M., 1998. Genetic structure of a natural
population of the ectomycorrhizal fungus Suillus pungens. New Phytol.
138, 533–542.
Biondini, M., 2001. A three-dimensional spatial model for plant competition in
a heterogenous soil environment. Ecol. Model. 142, 189–225.
Cline, E.T., Ammirati, J.F., Edmonds, R.L., 2005. Does proximity to mature
trees influence ectomycorrhizal fungus communities of Douglas-fir
seedlings? New Phytol. 166, 993–1009.
Cline, E., Vinyard, B., Edmonds, R., 2007. Spatial effects of retention trees on
mycorrhizas and biomass of Douglas-fir seedlings. Can. J. Forest Res. 37,
430–438.
de Montigny, L., 2004. Silviculture Treatments for Ecosystem Management in
the Sayward (STEMS): Establishment Report for STEMS 1, Snowden
Demonstration Forest Res. Br., BC Min. Forest, Victoria, BC Tech. Rep. 17.
Diamond, J.A., 1975. The island dilemma: lessons of modern biogeographic
studies for the design of natural reserves. Biol. Cons. 7, 129–146.
Dickie, I.A., Reich, P.B., 2005. Ectomycorrhizal fungal communities at forest
edges. J. Ecol. 93, 244–255.
Durall, D.M., Jones, M.D., Wright, E.F., Kroeger, P., Coates, K.D., 1999.
Species richness of ectomycorrhizal fungi in cutblocks in the ICH forests of
northwestern British Columbia. Can. J. Forest Res. 29, 1322–1332.
Author's personal copy
1352
M.D. Jones et al. / Forest Ecology and Management 255 (2008) 1342–1352
Gherbi, H., Delaruelle, C., Selosse, M.-A., Martin, F., 1999. High genetic
diversity in a population of the ectomycorrhizal basidiomycete Laccaria
amethystina in a 150-year-old beech forest. Mol. Ecol. 8, 2003–2013.
Goodman, D.M., Durall, D.M., Trofymow, J.A., 1996. Describing ectomycorrhizae. In: Goodman, D.M., Durall, D.M., Trofymow, J.A., Berch, S.M.
(Eds.), A Manual of Concise Descriptions of North American Ectomycorrhizae. Mycologue Publications and Canada—BC Forest Resource Development Agreement, Canadian Forest Service, Victoria, Canada, pp. 3A.1–
3A.5.
Goodman, D.M., Trofymow, J.A., 1996. Piloderma fallax (Libert) Stalpers + Pseudotsuga menziesii (Mirb.) Franco. CDE1. In: Goodman, D.M., Durall,
D.M., Trofymow, J.A., Berch, S.M. (Eds.), Concise descriptions of North
American ectomycorrhizae. Mycologue Publications, and Canada BC,
Forest Resource Development Agreement, Canadian Forest Service, Victoria, BC, pp. CDE1.1–CDE1.4.
Green, J., Bohannan, B.J.M., 2006. Spatial scaling of microbial biodiversity.
Trends Ecol. Evol. 21, 501–507.
Green, R.N., Klinka, K., 1994. A Field Guide for Site Identification and
Interpretation for the Vancouver Forest Region, Land Management Handbook, vol. 28. Province of British Columbia, Ministry of Forests, Research
Branch, 80 pp.
Hagerman, S.M., Jones, M.D., Bradfield, G.E., Gillespie, M., Durall, D.M.,
1999a. Effects of clear-cut logging on the diversity and persistence of
ectomycorrhizae at a subalpine forest. Can. J. Forest Res. 29, 124–134.
Hagerman, S.M., Jones, M.D., Bradfield, G.E., Sakakibara, S.M., 1999b.
Ectomycorrhizal colonization of Picea engelmannii ! Picea glauca seedlings planted across cut blocks of different sizes. Can. J. Forest Res. 29,
1856–1870.
Ingleby, K., Mason, P.A., Last, F.T., Fleming, L.V., 1990. Identification of
Ectomycorrhizas. ITE Research, De la Bastide and Kendrick Publication
No. 5. HMSO, London.
Izzo, A., Agbowo, J., Bruns, T.D., 2005. Detection of plot-level changes in
ectomycorrhizal communities across years in an old-growth mixed-conifer
forest. New Phytol. 166, 619–630.
Jones, M.D., Durall, D.M., Cairney, J.W.G., 2003. Ectomycorrhizal fungal
communities in young forest stands regenerating after clearcut logging.
New Phytol. 157, 399–422.
Jönsson, M.T., Fraver, S., Jonsson, B.G., Dynesius, M., Rydgård, M., Esseen, P.A., 2007. Eighteen years of tree mortality and structural changes in an
experimentally fragmented Norway spruce forest. Forest Ecol. Man. 242,
306–313.
Kernaghan, G., Berch, S.M., 1996. Lactarius scrobiculatus (Fr.) Fr. + Tsuga
heterophylla (Raf.) Sarg., CDE11. In: Goodman, D.M., Durall, D.M.,
Trofymow, J.A., Berch, S.M. (Eds.), Concise Descriptions of North
American Ectomycorrhizae. Mycologue Publications, and Canada—BC
Forest Resource Development Agreement, Canadian Forest Service,
Victoria, BC, pp. CDE11.1–CDE11.4.
Kranabetter, J.M., 1999. The effect of refuge trees on a paper birch ectomycorrhiza community. Can. J. Bot. 77, 1523–1528.
Kranabetter, J.M., Wylie, T., 1998. Ectomycorrhizal community structure
across forest openings on naturally regenerated western hemlock seedlings.
Can. J. Bot. 76, 189–196.
Luoma, D.L., Stockdale, C.A., Molina, R., Eberhart, J.L., 2006. The spatial
influence of Pseudotsuga menziesii retention trees on ectomycorrhiza
diversity. Can. J. Forest Res. 36, 2561–2573.
MacArthur, R.H., Wilson, E.O., 1967. The Theory of Island Biogeography.
Princeton University Press, Princeton.
Macdonald, S.E., Fenniak, T.E., 2007. Understory plant communities of boreal
mixedwood forests in western Canada: natural patterns and response to
variable-retention harvesting. Forest Ecol. Manage. 242, 34–48.
Meidinger, D., Pojar, J., 1991. Ecosystems of British Columbia. BC Ministry of
Forests, Victoria.
Outerbridge, R.A., Trofymow, J.A., 2004. Diversity of ectomycorrhizae on
experimentally planted Douglas-fir seedlings in variable retention forestry
sites on southern Vancouver Island. Can. J. Bot. 82, 1671–1681.
Parsons, W.F.J., Miller, S.L., Knight, D.H., 1994. Root-gap dynamics in a
lodgepole pine forest: ectomycorrhizal and nonmycorrhizal fine root
activity after experimental gap formation. Can. J. Forest Res. 24,
1531–1538.
Peay, K.G., Bruns, T.D., Kennedy, P.G., Bergemann, S.E., Garbelotto, M., 2007.
A strong species–area relationship for eukaryotic soil microbes: island size
matters for ectomycorrhizal fungi. Ecol. Lett. 10, 470–480.
Read, D.J., Perez-Moreno, J., 2003. Mycorrhizas and nutrient cycling in
ecosystems—a journey towards relevance? New Phytol. 157, 475–492.
Read, D.J., Leake, J.R., Perez-Moreno, J., 2004. Mycorrhizal fungi as drivers of
ecosystem processes in heathland and boreal forest biomes. Can. J. Bot. 82,
1243–1263.
Roxburgh, H., Shea, K., Wilson, J.B., 2004. The intermediate disturbance
hypothesis: patch dynamics and mechanisms of species coexistence. Ecology 85, 359–371.
Schimel, J.P., Bennett, J., 2004. Nitrogen mineralization: challenges of a
changing paradigm. Ecology 85, 591–602.
Siira-Pietikainen, A., Pietikainen, J., Fritze, H., Haimi, J., 2001. Short-term
responses of soil decomposer communities to forest management: clear
felling versus alternative forest harvesting methods. Can. J. Forest Res. 31,
88–99.
Simard, S.W., Perry, D.A., Smith, J.E., Molina, R., 1997. Effects of soil
trenching on occurrence of ectomycorrhizas on Pseudotsuga menziesii
seedlings grown in mature forests of Betula papyrifera and Pseudotsuga
menziesii. New Phytol. 136, 327–340.
Simberloff, D.S., 1972. Properties of the rarefaction diversity measurement.
Am. Nat. 106, 414–418.
Simberloff, D.S., Abele, L.G., 1976. Island biogeography theory and conservation practice. Science 191, 285–286.
Smith, J.E., Molina, R., Huso, M.M.P., Larsen, M.J., 2000. Occurrence of
Piloderma fallax in young, rotation-age, and old-growth stands of Douglasfir (Pseudotsuga menziesii) in the Cascade Range of Oregon, USA. Can. J.
Bot. 78, 995–1001.
Thompson, J.D., Gibson, T.J., Plewniak, F., Janmougin, F., Higgins, P.G., 1997.
The ClustalX Windows interface: flexible strategies for multiple sequence
alignment aided by quality control analysis tools. Nucleic Acids Res. 24,
4876–4882.
Twieg, B.D., Durall, D.M., Simard, S.W., 2007. Ectomycorrhizal fungal
succession in mixed temperate forests. New Phytol. 176, 437–447.