Content uploaded by Tesfaye Wubet
Author content
All content in this area was uploaded by Tesfaye Wubet
Content may be subject to copyright.
Tropical Ecology 51(2): 125-148, 2010 ISSN 0564-3295
© International Society for Tropical Ecology
www.tropecol.com
Species-rich but distinct arbuscular mycorrhizal communities in
reforestation plots on degraded pastures and in neighboring pristine
tropical mountain rain forest
INGEBORG HAUG1*, TESFAYE WUBET2, MICHAEL WEIß1, NIKOLAY AGUIRRE3,
MICHAEL WEBER4, SVEN GÜNTER4 & INGRID KOTTKE1
1Eberhard-Karls-Universität Tübingen, Institute of Evolution and Ecology,
Organismic Botany, Auf der Morgenstelle 1, D-72076 Tübingen, Germany
2Department of Soil Ecology, Helmholtz Centre for Environmental Research-UFZ,
Theodor-Lieser-Straße 4, D-06120 Halle-Saale, Germany
3Universidad Nacional de Loja, Forest Engineering, Ciudadela Universitaria
“La Argelia”; Casilla de correo: 11-01-249, Loja, Ecuador
4Institute of Silviculture, Department of Ecology, Technische Universität München,
Am Hochanger 13, D-85354 Freising
Abstract: For the first time in tropical mountain rain forest, arbuscular mycorrhizal fungal
richness and community composition was investigated from planted seedlings of Cedrela
montana, Heliocarpus americanus, Juglans neotropica and Tabebuia chrysantha in reforestation
plots on degraded pastures. A segment of fungal 18S rDNA was sequenced from the mycorr-
hizas. Sequences were compared with those obtained from mycorrhizas of adult trees of 30
species in the neighboring, pristine tropical mountain rain forest. In total, 193 glomeromycotan
sequences were analyzed, 130 of them previously unpublished. Members of Glomeraceae,
Acaulosporaceae, Gigasporaceae and Archaeosporales were found in both habitats, with Glomus
Group A sequences being by far the most diverse and abundant. Glomus Group A sequence type
richness did not appear to differ between the habitats; a large number was observed in both.
Glomus Group A sequence type composition, however, was found distinctly different. Seedlings
were rarely colonized by fungi of the pristine forest but trapped a number of fungi known from
other areas, which were rarely found in the pristine forest.
Resumen: Por vez primera en un bosque lluvioso tropical montano se investigó la riqueza
de hongos micorrícicos arbusculares y la composición de la comunidad en plántulas de Cedrela
montana, Heliocarpus americanus, Juglans neotropica y Tabebuia chrysantha, en parcelas de
reforestación ubicadas en pastizales degradados. Se hizo la secuenciación de un segmento del
ADNr fúngico de las micorrizas. Las secuencias fueron comparadas con las obtenidas de
micorrizas de árboles adultos de 30 especies en el vecino y prístino bosque tropical lluvioso
montano. En total se analizaron 193 secuencias de glomeromicotano, 130 de las cuales no
habían sido publicadas previamente. Se encontraron miembros de Glomeraceae, Acaulosporaceae,
Gigasporaceae y Archaeosporales en ambos hábitats, siendo las secuencias del Grupo A de
Glomus por mucho las más diversas y abundantes. La riqueza de tipos de la secuencia del
Grupo A de Glomus no pareció diferir entre hábitats; en ambos se observó un número mayor.
Sin embargo, se encontró que la composición tipo de la secuencia del Grupo A Glomus era
notablemente diferente. Las plántulas fueron colonizadas rara vez por hongos del bosque
prístino, pero atraparon un número de hongos conocidos de otras áreas, los cuales sólo fueron
hallados rara vez en el bosque prístino.
N
* Corresponding Author; e-mail: ingeborg.haug@uni-tuebingen.de
126 ARBUSCULAR MYCORRHIZAL COMMUNITIES
Resumo: Pela primeira vez na floresta tropical sempreverde de montanha, a riqueza dos
fungos micorrízicos arbusculares e a composição da comunidade foi investigada a partir de
plântulas de Cedrela montana, Heliocarpus americanus, Juglans neotropica e Tabebuia
chrysantha plantadas em parcelas reflorestadas em pastagens degradadas. Um segmento
fúngico 18SrDNA foi sequenciado a partir das micorrízas. As sequências foram comparadas com
as obtidas em micorrízas de árvores adultas de 30 espécies na floresta tropical primitiva
sempreverde de montanha vizinha. No total foram analisadas 193 glomeromycotan sequências,
130 delas não publicadas anteriormente. Nos dois habitats foram encontrados membros das
Glomeraceae, Acaulosporaceae, Gigasporaceae e Archaeosporales, sendo as sequências Glomus
grupo A de longe a mais diversa e abundante. A riqueza do tipo de sequência Glomus grupo A
não parece diferir entre os habitats; um grande número foi observado em ambas. Contudo, foi
encontrado que a composição da sequência do tipo Glomus grupo A foi distintamente diferente.
As plântulas eram raramente colonizadas pelos fungos da floresta primitiva mas fixaram fungos
conhecidas de outras áreas, as quais eram raramente encontradas na floresta primitiva.
Key words: Cedrela montana, degraded pastures, Glomeromycota, Heliocarpus
americanus, Juglans neotropica, neotropical mountain rain forest, reforestation, ribosomal
18S RNA gene, Setaria sphacelata, Tabebuia chrysantha.
Introduction
Arbuscular mycorrhizal fungi (AMF, Glomero-
mycota) are the main mycobionts in grasslands
and many tropical forests. While the historic use of
spore-based morphospecies has suggested that there
are relatively few species of AMF and that most
species are ubiquitous generalists, DNA sequen-
cing of AMF directly from mycorrhizas and the
study of their molecular taxonomy has revealed a
multitude of new and perhaps site-specific fungi
(Aldrich-Wolfe 2007; Husband et al. 2002; Kottke
et al. 2008; Wirsel 2004; Wubet et al. 2003, 2006a,
2006b). Sequence-based, comparative analysis of
root-colonizing arbuscular mycorrhizal fungal
communities in different ecosystems around the
globe indicated lower numbers of fungal taxa in
severely anthropogenically-altered habitats comp-
ared to tropical forests (Öpik et al. 2006). Loss of
AMF species was observed in temperate forests
converted to agriculture (Helgason et al. 1998) and
was assumed to be even more serious in tropical
lowland forests with a consequent negative impact
on reforestation (Alexander & Lee 2005; Janos
1996). Conversion of forests into pastures in the
tropics is typically accomplished by clear-felling
followed by burning (slash-and-burn). Burn
frequency and length of inter-fire intervals
influence the degree of disturbance, and long-term
repeated burning, according to the few published
results, reduces arbuscular mycorrhizal abundance
and diversity, especially in the upper few centi-
meters of soil (Bastias et al. 2006; Chen & Cairney
2002; Pattinson et al. 1999). However, the effect of
fire on soil fungal communities has so far been
unpredictable (see Cairney & Bastias 2007 for a
recent review) and no information was available on
such degraded, slowly recovering pastures in the
neotropical mountain forest areas. Tree seedlings
exposed to such degraded environments during
reforestation may struggle to find appropriate
mycobionts. Previous investigations revealed a
distinct and highly diverse AMF community in the
neotropical mountain rain forest (Kottke et al.
2008), but AMF potential of pastures neighboring
the forest on the opposite river side, selected for
reforestation with native tree species, was un-
known. As part of the reforestation trial (Weber et
al. 2008) we investigated, for the first time, AMF
richness and composition of planted seedlings of
four native tree species on three regeneration
stages of the pastures. We considered the seedlings
as “trap plants” in situ expecting that plants
growing under natural conditions would reflect the
mycorrhizal potential more accurately than plants
used in pot cultures in the greenhouse. We
compared AMF identity, richness and composition
trapped by the seedlings in reforestation plots with
our findings from adult trees in the neighboring,
species-rich forest. We are well aware of the bias
from comparing seedling associated fungi with
those from adult trees, but we were unable to
HAUG et al. 127
sample mycorrhizal fungi of seedlings in the forest,
since so few seedlings were observed on the dark
forest floor.
We hypothesized (1) that the trap plants in the
reforestation plots would be colonized by fewer
AMF taxa than the trees in the pristine forest, (2)
that the AMF community in the reforestation plots
would be distinct from that in the forest, (3) that at
least some of the forest AMF would be trapped by
the planted seedlings and (4) that fungi in the
pastures would be closely related to known AMF
with widespread distributions.
Here we present the initial results of
molecular phylogenetic analysis of AMF from the
mycorrhizas in reforestation plots within fire-
degraded pastures at a neotropical mountain site.
We compare these results with our current
knowledge about mycorrhizas in the neighboring
pristine forest and published data on other sites.
Materials and methods
Study sites
The study sites are located between 1800 and
2200 m above sea level on the slopes above the San
Francisco River, Cordillera Real, South Ecuador,
(3° 58’ S, 79° 4’ W). The tropical mountain rain
forest is preserved on the steep north-facing side of
the river, but was cleared for cattle pasture at
least 40 years ago using slash-and-burn techniques
on the less steep slopes of the south-facing river
bank. These pastures have been abandoned since
c.1990 (Makeschin et al. 2008). Information on the
plant composition of the forest, which is extra-
ordinary rich in tree species, is presented in
Homeier et al. (2008). Overviews on land-use
gradients (Beck et al. 2008), reforestation experi-
ments (Weber et al. 2008), climate (Bendix et al.
2008) and soils (Makeschin et al. 2008; Wilcke et
al. 2008) are also available. Details on the samp-
ling areas, recently abandoned pasture (R1),
abandoned pasture covered by bracken (R2), and
abandoned pastures covered by shrubs (R3) are
given in Appendix Table 1 (data from Aguirre
2007). The different types of regenerating pasture
were planted with six-month-old, nursery-raised
seedlings of Cedrela montana Moritz ex Turcz.,
Heliocarpus americanus L., Juglans neotropica
Diels and Tabebuia chrysantha (Jacq.) G. Nicholson,
all local species of the tropical mountain rain
forest.
Soils are similar on both sides of the San
Francisco River, with low amounts of P and N in
the mineral soil (Makeschin et al. 2008). Significant
effects of land use on soils were documented by
these authors. A strong initial loss of carbon after
burning, presumably accompanied by losses of N
and P, were found with the organic layer slowly
regenerating during succession. Generally, soils
are characterized by the accumulation of thick
organic layers (8-35 cm) on top of the mineral soil
in the pristine forest and large Ah horizons in the
pastures. C/N ratios were similar in both habitats,
but pH values were near four in the pristine forest
and near five in the pastures. Accordingly, exchan-
geable K, Ca and Mg levels were approxi-mately
four times higher, and Al, Fe, Mn levels were three
times lower, in the pastures than in the forest site
(Makeschin et al. 2008). Although plant-available
nutrients are considered to be very low, the
organic fraction of the soil is a large nutrient
reservoir that can be mobilized by mycorrhizal
fungi and other microbes (Wilcke et al. 2008).
Sampling
Sampling was carried out twice in the refore-
station plots in the degraded pastures, one year
and three years after planting. The survival rate of
seedlings two years after planting was 94 % for T.
chrysantha, 68 % for C. montana, 57 % for H.
americanus and 44 % for J. neotropica (Aguirre
2007). Roots were sampled by tracing of single
roots from the trunk from four to nine individuals
of C. montana, H. americanus, J. neotropica, T.
chrysantha and two individuals of the dominating
grass, Setaria sphacelata (Schumacher) Moss
(Table 1). Samples were also collected before
planting from seven nursery-raised seedlings in
total of C. montana, Cinchona officinalis L., H.
americanus, and Piptocoma discolor (Kunth.)
Pruski (Table 2) to obtain at least few data of the
mycorrhizal community in the nursery. More
seedlings were unfortunately not available for
sampling. Seedlings were raised in a mixture of
highland black soil, bed sand and forest humus
(2:1:1) in 560 cm³ polyethylene planting bags at a
nursery at the Universidad Nacional de Loja,
Ecuador. All substrates except forest humus were
steam-fumigated before planting.
Roots were cleaned under tap water the same
day they were harvested and degree of colonization
was determined using standard staining methods
(Haug et al. 2004). For each seedling from the
abandoned pastures or nursery a mix from the
excavated roots was used and five 1.5 ml tubes
were each filled with three fine roots, each 1 cm in
128 ARBUSCULAR MYCORRHIZAL COMMUNITIES
HAUG et al. 129
130 ARBUSCULAR MYCORRHIZAL COMMUNITIES
length (Tables 1 & 2). Tubes were installed on an
electric dryer at about 50 °C for 12 hours and roots
kept on silica gel for DNA isolation. This procedure
was necessary because of the high air humidity in
the tropics. Sampling in the forest was carried out
in established plots along an altitudinal gradient
comprising different forest types as described in
Haug et al. (2004). The same amount of rootlets
per individual tree was handled as described for
the pasture samples. Samples were collected from
30 different tree species and up to four individuals
per species (Table 3). Some of the sequences (33 %)
were published in Kottke et al. (2008), but subs-
tantial numbers of samples were newly sequenced
for this investigation (Table 3).
Processing of fungal DNA sequences
DNA was isolated from the dried root samples
using the DNAeasy Plant Mini Kit (Qiagen,
Hilden, Germany). One to five tubes were proce-
ssed per plant individual. The number of tubes
yielding sequences is given in Tables 1-3. For
phylogenetic analysis and molecular identification,
we sequenced sections of the fungal nuclear gene
coding for the small ribosomal subunit (18S;
nucSSU). The following primer combinations were
used in nested PCRs: Glomus Group A: first
SSU128/SSU1536IH, second SSU300/GLOM1310
rc; Glomus Group B: first SSU 817/NS8, second
SSU817/LETC1670rc; Acaulosporaceae: first SSU
817/NS8, second SSU817/ ACAU1660rc; Archaeo-
sporales: first SSU817/SSU1536IH, second SSU
817/ARCH1375rc. Details on the primers and nes-
ted PCRs are given in Appendix. The sequences
obtained were assigned to higher fungal groups
with BLAST searches (Altschul et al. 1997) using
the National Center for Biotechnology Information
(NCBI) database (http: //www.ncbi.nlm.nih.gov). We
checked for putative chimeric sequences (details in
Appendix). The glomeromycotan sequences were
deposited in GenBank; their accession numbers
are given in Tables 1-3. Accession numbers
starting with DQ (63 sequences) were published in
Kottke et al. (2008). Accession numbers starting
with EU (130 sequences) were deposited during
this study.
Phylogenetic analysis for identification of
fungi and sequence type definition of Glomus
Group A
BLAST hits with sequence similarities ≥ 99 %
were downloaded from GenBank and added to the
dataset. We also included sequences from AM taxa
identified from spores. Only one sequence was
included in the final tree when several inserts of a
cloned PCR product were very similar (1 to 5 bases
different) and appeared together in a terminal
cluster. Sequence alignments were done with
MAFFT (Katoh et al. 2005). We used PAUP*
version 4.0b10 (Swofford 2002) to estimate the
phylogenetic relationships of the sequences obtai-
ned. Neighbour-Joining analyses (Saitou & Nei
1987) using the BIONJ modification (Gascuel
1997) with Kimura 2-distances were carried out
and combined with bootstrap analyses (Felsenstein
1985) from 1000 replicates. Additionally, maxi-
mum-likelihood (ML) analysis using RAxML
(Stamatakis 2006) was done with Glomus Group A
sequences, with GTR+CAT as a DNA substitution
model for heuristic search and GTR+G for final
tree optimization, again combined with a bootstrap
analysis from 1000 replicates. Glomus Group A
sequences which showed sequence similarities ≥
99 % were defined as a sequence type. In Glomus
Group B, Archaeosporales and Gigasporaceae,
sequence types were not defined because inter-
specific differences were not observed due to high
sequence conservation within these groups. We
also did not define sequence types in the Acaulo-
sporaceae and Paraglomeraceae because only a few
sequences of known species are currently avai-
lable.
Evaluation of sequence-based diversity of
Glomeromycota from the pastures, the nursery
and the neighboring mountain rain forest
Presence or absence of individual Glomero-
mycota sequences were compiled in tables respe-
cting habitat, plant species and plant individuals
(Tables 1-3). The sequences and sequence types
obtained from the two samplings in the refore-
station plots were pooled because we did not
observe interpretable differences among the two
years.
Analysis of richness and community
composition of Glomus Group A sequence types
We calculated a sample-based rarefaction
accumulation curve with 95 % confidence intervals,
and estimated the total sequence type richness of
Glomus Group A with Chao 2 and Jackknife 2
using the software EstimateS, v.8.0.0 (Colwell
2006) set to “randomize samples without replace-
ment”.
Cluster analysis was performed on the
presence/absence data matrix of the AM fungal
HAUG et al. 131
w
132 ARBUSCULAR MYCORRHIZAL COMMUNITIES
HAUG et al. 133
sequence types of Glomus Group A comparing the
AM fungal communities of the two habitats. The
Ward clustering method with Squared Euclidean
distances was implemented in SPSS v 14. We also
carried out a chi-square test to determine whether
the AM fungal communities differed between the
two habitats. A Venn diagram was designed to
display Glomus Group A sequence types in refore-
station plots, nursery and pristine forest and to
label the sequence types known from other
investigations.
Results
Sequence-based composition of Glomeromycota
in reforestation plots in degraded pastures and
neighboring tropical mountain rain forest
In total, from both habitats and the nursery,
193 glomeromycotan sequences were obtained, 130
sequences published here for the first time (Tables
1-3). Fifty-six mycorrhizal samples of 29 individuals
from the reforestation plots yielded sequences of
Glomus Group A (63), Glomus Group B (2), Acaulo-
sporaceae (8), Archaeosporales (3), Gigaspo-raceae
(1) and Paraglomeracaeae (2) (Table 1). In the
mycorrhizas of one individual seedling, up to eight
glomeromycotan sequences were detected (Table 1).
Fifty-five mycorrhizal samples of 42 tree indi-
viduals from the pristine forest belonging to 23
genera from 18 families yielded sequences of
Glomus Group A (69), Glomus Group B (3), Acaulo-
sporaceae (16), Archaeosporales (11) and Gigas-
poraceae (3) (Table 3). In the mycorrhizas of one
individual tree, up to six glomeromycotan seque-
nces were detected (Table 3). Twelve mycor-rhizal
samples of seven individuals from the nursery
seedlings belonging to four genera yielded seque-
nces of Glomus Group A (5), Glomus Group B (4),
Acaulosporaceae (1) and Archaeosporales (2) (Ta-
ble 2).
Glomus Group A fungi were associated with
nearly all the plant individuals under investi-
gation, while Glomus Group B fungi were found
only with seedlings of J. neotropica and C.
montana from the reforestation plots, C. montana
and Piptocoma discolor from the nursery, and
three tree species in the pristine forest (Tables 1-3,
Appendix Fig. 1). Members of Acaulosporaceae
were detected in seedlings of C. montana and T.
chrysantha from the reforestation plots and P.
discolor from the nursery, but not in other
seedlings (Tables 1 & 2; Appendix Fig. 2). Acaulo-
sporaceae were also found associated with S. spha-
celata and with eleven tree species in the pristine
forest (Tables 1 & 3; Appendix Fig. 2). Members of
Archaeosporales were found in seedlings of T.
chrysantha and J. neotropica from the refore-
station plots (Table 1; Appendix Fig. 3), and C.
montana and H. americanus nursery plants (Table
2; Appendix Fig. 3). Archaeosporales seq-uences
were also obtained from eight tree species in the
pristine forest (Table 3; Appendix Fig. 3). Gigas-
poraceae were found only once on all the seedlings
(T. chrysantha) and on three tree species in the
pristine forest (Tables 1 & 3; Appendix Fig. 3).
Paraglomeraceae were only found on H. ameri-
canus and S. sphacelata on the reforestation plots
(Table 1; Appendix Fig. 3). An identical Gigas-
poraceae sequence (630 bp) was found with
Podocarpus oleifolius from pristine forest and T.
chrysantha from a reforestation plot (Appendix
Fig. 3). Two further Gigasporaceae sequences,
obtained from Faramea uniflora and Hyeronima
oblonga, are also very similar (Appendix Fig. 3).
Three of our sequences cluster with Archaeospora
trappei (Archaeosporaceae), two sequences of the
nursery form a sister clade to the Archaeo-
sporaceae. The rest of our sequences are in
clusters outside the known families of Archaeo-
sporales. Considerably more Archaeosporales seq-
uences were found in the forest than in the
reforestation plots (Appendix Fig. 3). The two
paraglomeracean sequences from H. americanus
and S. sphacelata form a separate cluster adjacent
to Paraglomus brasilianum and P. occultum (App-
endix Fig. 3).
Glomus Group A sequence type-based richness
and community composition
The Glomus Group A sequences (1035 bp) were
analyzed with BIONJ and ML. Both phylogenetic
trees showed very similar topologies and similar
bootstrap values (Fig. 1). Sequence types were
defined as sets of sequences with a sequence
similarity of ≥ 99 %. In most cases, sequence types
also formed monophyletic groups in our trees;
sequence types 3, 15 and 29 are not monophyletic
in the sequence trees. The 63 Glomus Group A
sequences from the reforestation plots were
grouped in 24 sequence types (Fig. 1; Table 4).
Nineteen sequence types are composed of two to
ten sequences; five sequence types consist of one
sequence only. All seedlings from the reforestation
plots shared fungal sequence types with other
seedlings and these sequence type clusters
consisted of sequences from different host species
134 ARBUSCULAR MYCORRHIZAL COMMUNITIES
Fig. 1. Phylogenetic relationships of Glomus Group A sequences obtained by use of primers SSU300 and
GLOM1310rc (in the second PCR) from mycorrhizas of tree seedlings in the reforestation plots (R1, R2, R3), in
a nursery (N), and trees in the pristine forest (F) in South Ecuador. Sequences from the study sites are
highlighted in bold. ML analysis was carried out on an alignment of nuclear DNA sequences coding for the
small ribosomal subunit (nucSSU; 1088 characters). The tree was rooted with Endogone pisiformis. Numbers
on branches designate bootstrap values (ML/BIONJ). Sequence types are based on sequence clusters with
sequence similarities ≥ 99 % (see text) and are numbered serially. Letters indicate occurrence of sequence types:
N-nursery, R-reforestation plots, F- pristine forest.
HAUG et al. 135
Fig. 1. Continued.
136 ARBUSCULAR MYCORRHIZAL COMMUNITIES
Fig. 2. Cluster analysis of Glomus Group A
sequence types: Cluster I- Sequence types of the
reforestation plots; Cluster II- Sequence types of the
nursery, refore-station and forest sites; Cluster III-
Sequence types of the reforestation and forest sites;
Cluster IV- Sequence types of the pristine forest site.
Fig. 3. Venn diagram of Glomus Group A sequence
types in the nursery, in the reforestion plots and in
the pristine forest. Black circles around black dots:
sequence type known as identified morphospecies or
showing ≥ 99 % similarity in the NS31-AM1 region
with published sequences from other studies (for
details see Table 4); circles shaded: sequence type
belongs to the most frequently detected AM-taxa
(Öpik et al. 2006).
(Fig. 1). Five sequence types from the reforestation
plots correspond to known morphospecies (st 1 =
Glomus intraradices, st 2 = G. vesiculiferum, st 3 =
G. fasciculatum, st 22 = G. proliferum, st 43 = G.
mosseae), nine are known as sequences from other
environmental studies and ten do not match any
currently published sequence (Table 4, Fig. 3).
Sequence types 11 and 24 were found on all
successional stages of the reforestation plots on
degraded pastures (R1, R2 and R3). Sequence type
24 was associated with all four planted tree species
and with S. sphacelata (Fig. 1). Sequence types 11
and 23 were verified for C. montana, J. neotropica
and T. chrysantha (Fig. 1).
The 69 Glomus Group A sequences from the
pristine forest (Table 3) were grouped in 25
sequence types (Fig. 1). Nineteen sequence types
are composed of two to ten sequences; six sequence
types consist of one sequence only. Nearly all
individual trees share fungal sequence types with
other trees, and sequence type clusters consist of
sequences from different tree species (Fig. 1). One
sequence type can be assigned to a morphospecies
(st 2 = Glomus vesiculiferum), five are known as
sequences from other studies and 19 do not match
any currently published sequence (Table 4; Fig. 3).
Seven Glomus Group A sequence types were
common in the reforestation plots and the forest;
three of these are known from other studies (Table
4; Fig. 3). Only one of the nursery sequence types
(st11) was found in the pristine forest. This
sequence type occurred also in the reforestation
plots and is thus the only sequence type found on
all three sites (Fig. 3). Three further sequence
types were common in the nursery and the
reforestation plots (Fig. 3). In the reforestation
plots, the known sequence types are high in
HAUG et al. 137
numbers (58 %), while in the pristine forest, the
proportion of new sequence types is high (76 %)
and that of known sequence types is low (24 %).
Richness estimation of Glomus Group A
sequence types from the reforestation site and the
pristine forest showed overlapping, thus not
significantly different accumulation curves at 95 %
confidence intervals. Accumulation curves did not
reach their asymptotes in either habitat (Appendix
Fig. 4). The expected richness (estimated with
Chao 2 and Jacknife 2) was slightly higher for the
reforestation site (Appendix Tables 4 & 5).
Ward cluster analysis was carried out based on
sequence type occurrence and separated four
distinct clades (Fig. 2). Cluster I contains 14 sequ-
ence types identified only from the reforestation plots.
Cluster IV is characterized by 18 sequence types
from the pristine forest only. Cluster II is com-
posed of five sequence types occurring in the nur-
sery, the reforestation plots and the forest site.
Cluster III is composed of six sequence types sha-
red between the reforestation plots and the forest
site. Habitat was found to have a significant
influence at P < 0.0001 (X2 = 200; df. = 84) on the
AM fungal composition.
Discussion
In this environmental study, we used mole-
cular methods for identifying arbuscular mycor-
rhizal fungi directly from the mycorrhizas. We
used SSU sequences because the amplification suc-
cess was satisfactory and SSU is the only gene
with a broad taxon sampling in Glomeromycota
(Redecker & Raab 2006). We found a number of
hitherto unknown AMF sequences in the refore-
station plots and the pristine forest. As seen in
other mycorrhizal community investigations (Hus-
band et al. 2002; Öpik et al. 2003; Scheublin et al.
2004; Whitfield et al. 2004), Glomus Group A
members dominated in both habitats. We cannot
exclude that choice of primers caused a bias for
this fungal group and future investigations may
reveal importance of other AMF in the area. Using
the primers currently available, we obtained
insufficient numbers of sequences to define phylo-
tyes in Glomus Group B, Acaulosporaceae, Gigas-
poraceae, Paraglomeraceae or Archaeosporales, and
thus we could not carry out a comparative analysis
on richness and community composition of these
fungal groups but confined these analysis to Glomus
Group A. We used sequence similarity of 99 % as a
criterion to create sequence types in Glomus Group
A. This high degree of similarity was nearly
always found within well-supported monophyletic
clades in our trees, corroborating our sequence
type definition.
We expected much lower numbers of AMF in
the reforestation plots according to previous
findings at severely anthropogenically influenced
sites (Alexander et al. 1992; Cairney & Bastias
2007; Janos 1996; Öpik et al. 2006). However,
AMF from nearly all the taxonomic groups were
present and fungal richness of Glomus Group A in
the reforestation plots was equal to the richness in
the neighboring forest and a tropical forest of
Panama (Husband et al. 2002). Arbuscular mycor-
rhizal fungal potential in the reforestation plots on
the degraded pastures in the tropical mountain
forest area apparently differs from
Table 4. Occurrence of Glomus Group A sequence types in the habitats (N = nursery, R = reforestation plots on
degraded pastures, F = pristine forest), number of sequences per habitat, and identification of sequence types
by comparison with morpho-species or sequences in the SSU300-GLOM1310 or in the NS31-AM1 region from
other investigations.
Sequence
type
Number of sequences
N R F
Sequence type known as
identified morpho-species or
as sequence in the SSU300-
Glom1310 region
Sequence type known as identified
morpho-species or showing ≥ 99 %
similarity with published sequences in
the NS31-AM1 region (with host/
isolation source)
1 1 3 0 Glomus intraradices Glomus intraradices
2 0 2 2 Glomus vesiculiferum Glomus vesiculiferum
3 1 3 0 Glomus fasciculatum Glomus fasciculatum
Contd...
138 ARBUSCULAR MYCORRHIZAL COMMUNITIES
Table 4. Continued.
Sequence
type
Number of sequences
N R F
Sequence type known as
identified morpho-species or
as sequence in the SSU300-
Glom1310 region
Sequence type known as identified
morpho-species or showing ≥ 99 %
similarity with published sequences in
the NS31-AM1 region (with host/
isolation source)
4 0 1 2 EU350063 Caragana korshinskii
EU332708 Glycine max
EF041068 Agrostis stolonifera
DQ357107 Ammophila arenaria
AJ563882 Phragmites australis
AY70 2066 grass roots
AM746139 soil
5 0 1 2 DQ357081 Ammophila arenaria
EU350066 Caragana korshinskii
6 0 4 0 AM412080 root tissue
AY129603 Glo35 Faramea occidentalis,
Tetragastris panamensis
7 0 0 1
8 1 0 0
9 0 0 2 EU417619 Afrothismia winkleri
10 0 0 1
11 1 7 2 (Glomus sinuosum 98 %) (Glomus sinuosum 98 %)
12 0 0 2
13 0 1 0
14 0 0 4
15 0 0 6
16 0 1 2
17 0 1 0 AB183987 roots in forest
18 0 0 2 AB183953 roots in forest
19 0 0 1
20 0 5 0
21 0 3 0 AJ699068 Marchantia foliacea
AF485887 Glo3a
Glomus sp. UY1225
AJ699068 Marchantia foliacea
22 0 2 0 Glomus proliferum Glomus proliferum
23 0 4 0
24 0 8 1 AJ854084 Glo2 Ajuga reptans
AY969156 mixed hardwood soil
25 0 1 0
26 0 0 2
27 0 0 2
28 0 0 2
29 0 0 2
30 0 0 1
31 0 2 5
32 0 3 0 DQ085211 Juniperus procera DQ085211 Juniperus procera
Contd...
HAUG et al. 139
Table 4. Continued.
Sequence
type
Number of sequences
N R F
Sequence type known as
identified morpho-species or
as sequence in the SSU300-
Glom1310 region
Sequence type known as identified
morpho-species or showing ≥ 99 %
similarity with published sequences in
the NS31-AM1 region (with host/
isolation source)
33 0 0 3
34 0 0 3
35 0 0 1
36 0 0 1
37 0 1 0
38 0 1 0
39 0 5 0 AJ699069 Marchantia foliacea
AJ563913 Phragmites australis
AM746141 soil
AJ854089 Glo18 Ajuga reptans
40 0 2 0 EF041100 Glo60 Agrostis
stolonifera
AM746145 soil
41 0 0 10 AY129609 Glo32 Faramea
occidentalis
42 0 1 0 DQ085256 Juniperus procera DQ085254 Juniperus procera
AM849309 Hepatica nobilis
43 1 1 0 Glomus mosseae Glomus mosseae
that found in other heavily anthropogenically-
influenced sites. The man-made tropical mountain
pastures may better be categorized as “grassland”
according to the classification suggested by Öpik et
al. (2006). These authors found no significant
differences between tropical forests and grasslands
in the average number of arbuscular mycorrhizal
taxa per plant species. Highly diverse secondary
vegetation is found in the surrounding areas of the
abandoned pastures (Martinez et al. 2008) and
may supply AMF propagules to the reforestation
plots. Plant species richness was experimentally
shown to support AMF diversity (van der Heijden
et al. 1998). Investigation of the AMF community
of the secondary woody vegetation in the surrou-
ndings is, however, needed to corroborate our
assumption. Glomus A sequences from mycor-
rhizas of the dominating grass species S. spha-
celata clustered with sequences from seedlings
indicating that propagules are transferred from
the pasture vegetation to the seedlings. In cont-
rast, Aldrich-Wolfe (2007) found minimal sharing
of mycobionts between tree seedlings and pasture
grasses.
Our expectation of distinct fungal composition
in the habitats was corroborated by the Ward
cluster analysis on presence/absence data of Glomus
Group A sequence types. Distinct differences in
fungal composition resulted also from a global
overview on molecular defined AMF by Öpik et al.
(2006). Most of the Glomus Group A sequence
types were found either in the reforestation plots
or in the pristine forest. Only a small number of
fungal sequence types occurred in both habitats.
Thus, the seedlings, during the first three years,
trapped only a very limited number of fungi so far
known from the pristine neighboring forest.
Husband et al. (2002) found a change of AMF
associations during successive field studies on
tropical tree seedlings, indicating that plant age
could influence the fungal community. Only further
observations of the seedlings and comparative
investigation of AMF of remnant trees and other
woody plants in the surroundings of the refore-
station plots can show if forest adapted fungi are
still present or were definitely lost during deg-
radation of the habitat. Such future investigations
would also contribute to lowering the potential
bias of only investigating tree seedlings for AMF
potential of a site. Seedlings may be preferentially
colonized by fast growing AMF and thus, we may
have missed some of forest adapted AMF on the
140 ARBUSCULAR MYCORRHIZAL COMMUNITIES
degraded pastures.
We did not observe differences in sequence
type presence among the three succession stages of
the reforestation plots on the degraded pastures or
along the altitudinal gradient in the pristine
forest. The database may still be too limited to
detect such differences. The occurrence of only one
sequence type common between pristine forest and
nursery may indicate difficulties in cultivating
forest AMF in the nursery. However, this result
needs to be corroborated by further sampling. The
occurrence of widespread fungi like Glomus
intraradices, G. fasciculatum and G. mosseae in
the nursery seedlings was most likely due to
inefficient soil sterilization.
Our results indicate strong habitat influence
on AMF community, corresponding to conclusions
by Öpik et al. (2006) in their global analysis. Al-
though the two habitats generally share the same
climate - nearly permanent high precipitation and
moderate temperature - roots in the upper few cen-
timeters of the pasture soils in the reforestation
plots are more frequently stressed by water defi-
ciency and high temperature from intense solar
radiation. Even more importantly, the differences
in soil nutrients and the occurrence of a large Ah
horizon on the reforestation plots versus a thick
humus layer in the pristine forest may influence
the AMF communities. Some of the fungi may
have been imported by human activities from
other continents and are now spread by S.
sphacelata, as this grass species is predominantly
planted by local people to establish the pastures
after burning.
Of the 13 most frequently detected AM fungal
taxa (Öpik et al. 2006, Table 2) the following are
found on the reforestation plots: Glomus intra-
radices/fasciculatum (st 1, st 3), G. mosseae (=st
43), Glomus sp. UY1225 (=st 21), Glo18 (=st 39)
and Glo2 (=st 24). In the pristine forest only one
(Glo2 = st 24) of these globally occurring AM
fungal taxa was detected (Fig. 3). All these taxa
are generalists colonizing a wide host range
(Helgason et al. 2007), which is enlarged with
every new evaluation (this study; Liu et al. 2009;
Öpik et al. 2008). The occurrence on the degraded
pastures confirms the better resilience of these
taxa (Helgason et al. 2007). Concerning these
comparisons one has to keep in mind, however,
that several genotypes/strains/subspecies may be
hidden behind (Croll et al. 2008; Mathimaran et al.
2008). No host-specific fungal sequence types are
obvious from the data of the reforestation plots
supporting the above mentioned observations and
Öpik et al. (2006). However, host-fungus prefe-
rences, as found in studies of semi-dry tropical
forest (Wubet et al. 2006a, 2006b) or in legumes
and non-legumes (Scheublin et al. 2004) may be
obscured in our investigation because of still
insufficient sampling.
Acknowledgements
We wish to thank Prof. F. Oberwinkler for his
long-term support, Dr. J. Homeier for identification
of the trees in the pristine forest, Jutta Bloschies
for assistance in DNA sequencing and Laura
Aldrich-Wolfe for critically reading the manuscript.
The research was financially supported by DFG
FOR402.
References
Aguirre, N.M. 2007. Silvicultural contributions to the
reforestation with native species in the tropical
mountain rain forest region of South Ecuador. Ph.D.
Thesis. Technische Universität München, Germany.
Aldrich-Wolfe, L. 2007. Distinct mycorrhizal commu-
nities on new and established hosts in a transitional
tropical plant community. Ecology 88: 559-566.
Alexander, I.J., N. Ahmad & S.S. Lee. 1992. The role of
mycorrhizae in the regeneration of some Malaysian
forest trees. Philosophical Transactions of the Royal
Society B, 335: 379-388.
Alexander, I.J. & S.S. Lee. 2005. Mycorrhizas and
ecosystem processes in tropical rain forest: impli-
cations for diversity. pp. 165-203. In: D.F. Burslem,
M.A. Pinard & S.E. Hartley (eds.) Biotic Interactions
in the Tropics: Their role in the Maintenance of
Species Diversity. Cambridge University Press,
Cambridge.
Altschul, S.F., T.L. Madden, A.A. Schäffer, J. Zhang, Z.
Zhang, W. Miller & D.J. Lipman. 1997. Gapped
BLAST and PSI-BLAST: a new generation of
protein database search programs. Nucleic Acids
Research 25: 3389-3402.
Bastias, B.A., Z.Q. Huang, T. Blumfield, Z. Xu & J.W.
Cairney. 2006. Influence of repeated prescribed
burning on the soil fungal community in an eastern
Australian wet sclerophyll forest. Soil Biology and
Biochemistry 38: 3492-3501.
Beck, E., J. Bendix, I. Kottke, F. Makeschin & R.
Mosandl (eds.) 2008. Gradients in a Tropical
Mountain Ecosystem of Ecuador. Series Ecological
Studies198, Springer Verlag, Berlin, Heidelberg.
Bendix, J., R. Rollenbeck, M. Richter, P. Fabian & P.
Emck. 2008. Climate. pp. 63-74. In: E. Beck, J.
Bendix, I. Kottke, F. Makeschin & R. Mosandl (eds.)
Gradients in a Tropical Mountain Ecosystem of
HAUG et al. 141
Ecuador. Series Ecological Studies 198, Springer
Verlag, Berlin, Heidelberg.
Cairney, J.W. G. & B.A. Bastias. 2007. Influences of fire
on forest soil fungal communities. Canadian Jou-
rnal Forest Reserach 37: 207-215.
Chen, D.M. & J.W. Cairney. 2002. Investigations on the
influence on prescribed burning on the ITS profiles
of ectomycorrhizal and other soil fungi at three
Australian sclerophyll forest sites. Mycological Rese-
arch 106: 533-540.
Colwell, R.K. 2006. EstimateS: Statistical estimation of
species richness and shared species from samples.
Version 8. Persistent URL <purl.oclc.org/estimates>.
Croll D., L. Wille, H.A. Gamper, N. Mathimaran, P.J.
Lammers, N. Corradi & I.R. Sanders. 2008. Genetic
diversity and host plant preferences revealed by
simple sequence repeat and mitochondrial markers
in a population of the arbuscular mycorrhizal fun-
gus Glomus intraradices. New Phytologist 178: 672-
687.
Felsenstein, J. 1985. Confidence limits on phylogenies:
an approach using the bootstrap. Evolution 39:
83-791.
Gascuel, O. 1997. BIONJ: An improved version of the NJ
algorithm based on a simple model of sequence data.
Molecular Biology and Evolution 14: 685-695.
Haug, I., J. Lempe, J. Homeier, M. Weiß, S. Setaro, F.
Oberwinkler & I. Kottke. 2004. Graffenrieda emar-
ginata (Melastomataceae) forms mycorrhizas with
Glomeromycota and with a member of the Hymeno-
scyphus ericae aggregate in the organic soil of a
neotropical mountain rain forest. Canadian Journal
of Botany 82: 340 356.
Helgason, T., T.J. Daniell, R. Husband, A.H. Fitter &
J.P. Young. 1998. Ploughing up the wood-wide web?
Nature 394: 431.
Helgason, T., J.W. Merryweather, J.P.W. Young & A.H.
Fitter. 2007. Specifity and resilience in the
arbuscular mycorrhizal fungi of a natural woodland
community. Journal of Ecology 95: 623-630.
Homeier, J., F.A. Werner, S.R. Gradstein, S.-W. Breckle,
& M. Richter. 2008. Potential vegetation and
floristic composition of Andean forests in South
Ecuador, with a focus on the RBSF. Ecological
Studies 198: 87-100.
Husband, R., E.A. Herre, S.L. Turner, R. Gallery, & J. P.
W. Young. 2002. Molecular diversity of arbuscular
mycorrhizal fungi and patterns of host association
over time and space in a tropical forest. Molecular
Ecology 11: 2669-2678.
Janos, D.P. 1996. Mycorrhizas, succession, and the
rehabilitation of deforested lands in the humid
tropics. pp. 129-162. In: J.C. Frankland, N. Magan
& G.M. Gadd (eds.) Fungi and Environmental
Change. Cambridge University Press, Cambridge.
Katoh, K., K. Kuma, H. Toh & T. Miyata. 2005. MAFFT
version 5: improvement in accuracy of multiple
sequence alignment. Nucleic Acids Research 33:
511-518.
Kottke, I. 2002. Mycorrhizae - Rhizosphere determinants of
plant communities. pp. 919-932. In: Y. Waisel, A.
Eshel & U. Kafkafi (eds.) Plant Roots: The Hidden
Half. Marcel Dekker Incorporation, New York, Ba-
sel.
Kottke, I., I. Haug, S. Setaro, J.P. Suárez, M. Weiß, M.
Preußing, M. Nebel, M. & F. Oberwinkler. 2008.
Guilds of mycorrhizal fungi and their relation to
trees, ericads, orchids and liverworts in a neo-
tropical mountain rain forest. Basic and Applied
Ecology 9: 13-23.
Liu Y., L. Hei, L. An, T. Helgason & H. Feng. 2009.
Arbuscular mycorrhizal dynamics in a chrono-
sequence of Caragana korshinskii plantations.
FEMS Microbiology Ecology 67: 81-92.
Makeschin, F., F. Haubrich, M. Abiy, J.I. Burneao & T.
Klinger. 2008. Pasture management and natural
soil regeneration. pp. 397-408. In: E. Beck, J. Ben-
dix, I. Kottke, F. Makeschin & R. Mosandl (eds.)
Gradients in a Tropical Mountain Ecosystem of
Ecuador. Series Ecological Studies 198, Springer
Verlag, Berlin, Heidelberg.
Martinez, A., M.D. Mahecha, G. Lischeid, & E. Beck.
2008. Succession stages of vegetation regeneration:
secondary tropical mountain forest. pp. 409-415. In:
E. Beck, J. Bendix, I. Kottke, F. Makeschin & R.
Mosandl (eds.) Gradients in a Tropical Mountain
Ecosystem of Ecuador. Series Ecological Studies
198, Springer Verlag, Berlin, Heidelberg.
Mathimaran, N., L. Falquet, K. Ineichen, C. Picard, D.
Redecker, T. Boller & A. Wiemken. 2008. Micro-
satellites for disentangling underground networks:
Strain specific identification of Glomus intraradices,
an arbuscular mycorrhizal fungus. Fungal Genetics
and Biology 45: 812-817.
Öpik, M., M. Moora, J. Liira, U. Koljalg, M. Zobel & R.
Sen. 2003. Divergent arbuscular mycorrhizal fungal
communities colonize roots of Pulsatilla spp. in
boreal Scots pine forest and grassland soils. New
Phytologist 160: 581-593.
Öpik, M., M. Moora, J. Liira & M. Zobel. 2006.
Composition of root-colonizing arbuscular myco-
rrhizal fungal communities in different ecosystems
around the globe. Journal of Ecology 94: 778-790.
Öpik, M., M. Moora, M. Zobel, Ü. Saks, R. Wheatley, F.
Wright & T. Daniell. 2008. High diversity of
arbuscular mycorrhizal fungi in a boreal herb-rich
coniferous forest. New Phytologist 179: 867-876.
142 ARBUSCULAR MYCORRHIZAL COMMUNITIES
Pattinson, G.S., K.A. Hammill, B.G. Sutton & P.A.
McGee. 1999. Simulated fire reduces the density of
arbuscular mycorrhizal fungi at the soil surface.
Mycological Reserach 103: 491-496.
Redecker, D. & P. Raab. 2006. Phylogeny of the Glome-
romycota (arbuscular mycorrhizal fungi): recent
developments and new gene markers. Mycologia 98:
885-895.
Saitou, N. & M. Nei. 1987. The neighbour-joining method:
a new method for reconstructing phylogenetic trees.
Molecular Biology and Evolution 4: 406-425.
Scheublin, T.R., K.P. Ridgway, J.P.W. Young & M.A.G.
van der Heijden. 2004. Nonlegumes, legumes, and
root nodules harbor different arbuscular mycorrhizal
fungal communities. Applied and Environmental
Microbiology 70: 6240-6246.
Smith, S.E. & D.J. Read. 2008. Mycorrhizal Symbiosis.
3rd edn., Academic Press, London.
Stamatakis, A. 2006. RAxML-VI-HPC: maximum likeli-
hood-based phylogenetic analyses with thousands of
taxa and mixed models. Bioinformatics 22: 2688-
2690.
Swofford, D.L. (2002). PAUP*. Phylogenetic Analysis
Using Parsimony (*and other Methods), version
4b10. Sinauer Associates, Sunderland, MA, USA.
Van der Heijden, M.G., J.N. Klironomos, M. Ursic, P.
Moutoglis, R. Streitwolf-Engel, T. Boller, A. Wiem-
ken & I.R. Sanders. 1998. Mycorrhizal fungal diver-
sity determines plant biodiversity, ecosystem varia-
bility and productivity. Nature 396: 69-72.
Weber, M., S. Günter, N. Aguirre, B. Stimm & R.
Mosandl. 2008. Reforestation of abandoned pastures:
Silvicultural means to accelerate forest recovery and
biodiversity. pp. 431-441. In: E. Beck, J. Bendix, I.
Kottke, F. Makeschin, & R. Mosandl (eds.)
Gradients in a Tropical Mountain Ecosystem of
Ecuador. Series Ecological Studies 198, Springer
Verlag, Berlin, Heidelberg.
Whitfield, L., A.J. Richards & D.L. Rimmer. 2004.
Relationships between soil heavy metal concentration
and mycorrhizal colonization in Thymus polytrichus
in northern England. Mycorrhiza 14: 55-62.
Wilcke, W., S. Yasin, A. Schmitt, C. Valarezo & W. Zech.
2008. Soils along the altitudinal transect and in
catchments. pp. 75-85. In: E. Beck, J. Bendix, I.
Kottke, F. Makeschin, & R. Mosandl (eds.) Gradients
in a Tropical Mountain Ecosystem of Ecuador.
Series Ecological Studies 198, Springer Verlag,
Berlin, Heidelberg.
Wirsel, S.G.R. 2004. Homogenous stands of a wetland
grass harbour diverse consortia of arbuscular
mycorrhiza. FEMS Microbiology Ecology 48: 129-
138.
Wubet, T., M. Weiß, I. Kottke, D. Teketay & F. Ober-
winkler. 2003. Molecular diversity of arbuscular
mycorrhizal fungi in Prunus africana, an endangered
medicinal tree species in dry Afromontane forests of
Ethiopia. New Phytologist 161: 517-528.
Wubet, T., M. Weiß, I. Kottke & F. Oberwinker. 2006a.
Two threatened coexisting indigenous conifer
species in the dry Afromontane forests of Ethiopia
are associated with distinct arbuscular mycorrhizal
fungal communities. Canadian Journal of Botany
84: 1617-1627.
Wubet, T., M. Weiß, I. Kottke, D. Teketay & F.
Oberwinker. 2006b. Phylogenetic analysis of nuclear
small subunit rDNA sequences suggests that the
endangered African Pencil Cedar, Juniperus
procera, is associated with distinct members of
Glomeraceae. Mycological Research 110: 1059-1069.
Appendix
Detailed methods for molecular identification
of arbuscular mycorrhizal fungi
We used SSU817 as a forward primer with
GLOM1310rc primer in a first attempt, by which
we obtained 127 sequences of Glomus Group A
(data not shown). However, because of the high
similarity of about 550 nucleotides, the phylogenetic
trees were poorly resolved. To improve the phylo-
genetic resolution by using longer sequences, we
used SSU128 or SSU300 as forward primers in the
first PCR in subsequent attempts (Tables 2 & 3).
We also tried to use the SSU300 forward primer
with the reverse primers LETC1670rc, ACAU1660rc
and ARCH1375rc, but no PCR products were obtai-
ned. Amplification with the primers ACAU 1660rc,
LETC1670rc, ARCH1375rc, GIGA5.8R was conspi-
cuously less successful than with GLOM 1310rc.
Even an amplification success did not guarantee
that the sequence belonged to the Acaulospo-
raceae, Glomus Group B, Archaespor-ales or Giga-
sporaceae, respectively.
The PCR reaction volume was 50 μl, with
concentrations of 3 mM MgCl2, 200 μM of each
dNTP (Life Technologies, Eggenstein, Germany),
0.5 μM of each of the primers (Biomers, Ulm,
Germany), 1 U Taq Polymerase (Life Technologies),
amplification buffer (Life Technologies), 0.2 μl 1 %
BSA (bovine serum albumin; Sigma) and 1 μl DNA
extract in the first PCR, and 0.5 μl of the first PCR
products for the second PCR. Three microliters of
each reaction were separated on a 1.5 % agarose
gel and stained with ethidium bromide prior to
direct sequencing or cloning. Amplified PCR
products were cloned with the Invitrogen TA Cloning
HAUG et al. 143
Kit (Life Technologies) following the manufac-
turer’s instructions. Inserts were reamplified from
clones using the M13 primers by picking twelve
positive bacterial clones with a toothpick and
placing them directly into the PCR reaction
mixture. After gel electrophoresis up to 12 positive
cloned amplification products were cleaned with
QIAquick (Qiagen, Hilden, Germany). Direct seq-
uencing of PCR products was performed using the
forward PCR primer as the sequencing primer; for
cloned products the M13F primer was used. After
preliminary analysis, the second strand was
sequenced with the reverse PCR primer M13R. Cycle
sequencing was conducted using the ABI PRISM
Dye-Terminator Cycle Sequencing Kit (Applied
Biosystems, Foster City, CA, USA) according to
the manufacturer’s protocol, but with the reaction
volume halved and the kit diluted 1:6. Electro-
phoresis and data sampling were performed on an
automated sequencer (ABI 3100; Applied Bio-
systems). Sequences were edited and contigs cons-
tructed using Sequencher software (version 4.1,
Gene Codes, Ann Arbor, Michigan). The sequences
obtained were assigned to higher fungal groups
with BLAST searches (Altschul et al. 1997) against
the National Center for Biotechnology Information
(NCBI) database (http://www.ncbi. nlm.nih.gov). We
checked for putative chimeric sequences using the
program Pintail (Ashelford et al. 2005) and we
compared sequence segments with GenBank acce-
ssions using a BLAST search. About 16 % of seque-
nced clones had portions of sequences that matched
other taxa (plants, ascomycetes, basidiomycetes).
These sequences were excluded.
Appendix Table 1. The characteristics of the three reforestation plots (source: Aguirre 2007).
Characteristics R1: Pasture R2: Fern R3: Shrub
Altitude (m a.s.l.) 1800 - 2100 1850 - 2100 2000 - 2200
Inclination (%) 53 (6-90) 69 (10-100) 44 (5-55)
Vegetation cover (%) 100 100 80 - 100
Dominant life-forms grasses fern and few shrubs shrubs, fern and herbs
Actual use before
planting
livestock farming (cattle
pasture)
early successional state
dominated by fern
advanced successional state
dominated by shrubs
Shannon-Index 0.87 (0.20 - 1.380) 0.84 (0.58- 1.26) 1.89 (0.87 - 3.00)
Topography irregular and steep irregular and steep irregular and steep
Remnant trees Piptocoma discolor, Isertia
laevis, Tabebuia chrysantha
Nectandra membranacea,
Inga sp.
Vismia ferrruginea,
Tabebuia chrysantha,
Clethra sp.
Appendix Table 2. Primer names, sequences and references.
Primer Name Sequence (5´-3´) Reference
SSU128 GGA TAA CCG TGG TAA TTC TAG designed for this study
SSU1536IH RTT GYA ATG CYC TAT CCC CA Borneman & Hartin 2000, modified
SSU300 CAT TCA AAT TTC TGC CCT ATC A designed for this study
GLOM1310rc TAA CAA TGT TAG RCC TAG CT Redecker 2000
ACAU1660rc CCG ATC CGA GAG TCT CA Redecker 2000
LETC1670rc ACT CAC CGA TCG CCG ATC Redecker 2000
ARCH1375rc TCA AAC TTC CGT TGG CTA RTC GCR C Russell et al. 2002
NS8 TCC GCA GGT TCA CCT ACG GA White et al. 1990
NS5 AAC TTA AAG GAA TTG ACG GAA G White et al. 1990
ITS4 TCC TCC GCT TAT TGA TAT GC White et al. 1990
GIGA5.8R ACT GAC CCT CAA GCA KGT G Redecker 2000
GIGAIH CCC ATC ACG ATG AAR TTT CA designed for this study
144 ARBUSCULAR MYCORRHIZAL COMMUNITIES
Appendix Table 3. Primer combinations and annealing temperatures used for the PCRs.
Fungal group 1st PCR 2nd PCR Annealing temp. (°C)
Glomus Group A SSU817-SSU1536IH
SSU128-SSU1536IH
SSU817-GLOM1310rc
SSU128-GLOM1310rc
50
50
SSU128-SSU1536IH SSU300-GLOM1310rc 50
Glomus Group B SSU817-NS8 SSU817-LETC1670rc 50
Acaulosporaceae SSU817-NS8 SSU817-ACAU1660rc 50
Archaeosporales SSU817-SSU1536IH SSU817-ARCH1375rc 50
Gigasporaceae NS5-ITS4 NS5-GIGA5.8R 50
SSU817-NS8 SSU817-GIGAIH 50
Appendix Table 4. Sequence type richness esti-
mation in the pristine forest (n number of analyzed
PCR products, Sobs observed richness obtained by
resampling without replacement).
n Sobs Chao 2 Jackknife 2
10 6.00 37.06 15.44
25 20.32 30.85 35.29
42 25.00 27.42 31.91
Appendix Table 5. Sequence type richness esti-
mation in the reforestation plots (n number of
analyzed PCR products, Sobs observed richness
obtained by resampling without replacement).
n Sobs Chao 2 Jackknife 2
10 9.24 24.89 13.79
25 16.55 27.71 29.01
42 20.97 29.89 34.08
57 24.00 30.82 37.59
HAUG et al. 145
Appendix Figure 1. Glomus Group A and Glomus Group B sequences obtained from mycorrhizae of tree
species on reforestation plots (R1, R2, R3), pristine forest (F) and a nursery (N) in South Ecuador by use of
primers SSU817 - GLOM1310rc or SSU817 - LETC1670rc respectively (in the second PCR). BIONJ analysis
was carried out from an alignment of nuclear DNA sequences coding for the small ribosomal subunit (nucSSU;
912 characters). The tree was rooted with Endogone pisiformis. Numbers on branches designate bootstrap
values.
146 ARBUSCULAR MYCORRHIZAL COMMUNITIES
Appendix Figure 2. Acaulosporaceae sequences obtained by use of primers SSU817 and ACAU1660rc (in the
second PCR) from tree species mycorrhizas sampled in rehabilitation plots (R1, R2, R3), pristine forest (F) and
a nursery (N) in South Ecuador. BIONJ analysis was carried out from an alignment of nuclear DNA sequences
coding for the small ribosomal subunit (nucSSU; 900 characters). The tree was rooted with four sequences of
the Diversisporaceae. Numbers on branches designate bootstrap values.
Twelve sequences of Acaulosporaceae from the pristine forest cluster together. Each investigated Graffenrieda
emarginata individual showed an Acaulosporaceae sequence (Table 3).
HAUG et al. 147
Appendix Figure 3. Archaeosporales, Gigasporaceae and Paraglomeraceae sequences obtained by use of
primers SSU817 and ARCH1375rc, ACAU1660rc, LETC1670rc (in the second PCR) from mycorrhizas of tree
species on reforestation plots (R1, R2, R3), pristine forest (F) and a nursery (N) in South Ecuador. BIONJ
analysis was carried out from an alignment of nuclear DNA sequences coding for the small ribosomal subunit
(nucSSU; 631 characters). The tree was rooted with Endogone pisiformis. Numbers on branches designate
bootstrap values.
148 ARBUSCULAR MYCORRHIZAL COMMUNITIES
Appendix Figure 4. Sequence type accumulation curves with 95 % confidence intervals for the reforestation
plots and pristine forest. Sobs (number of sequence types observed by resampling without replacement).
Appendix References
Aguirre, N.M. 2007. Silvicultural contributions to the
reforestation with native species in the tropical
mountain rainforest region of South Ecuador. Ph.D.
Thesis. Technische Universität München, Germany.
Altschul, S.F., T.L. Madden, A.A. Schäffer, J. Zhang, Z.
Zhang, W. Miller & D.J. Lipman. 1997. Gapped
BLAST and PSI-Blast: a new generation of protein
database search programs. Nucleic Acids Research
25: 3389-3402.
Ashelford, K.E., N.A. Chuzhanova, J.C. Fry, A.J. Jones
& A.J. Weightman. 2005. At least 1 in 20 16S rRNA
sequence records currently held in public reposi-
tories is estimated to contain substantial anomalies.
Applied and Environmental Micro-biology 71: 7724-
7736.
Borneman, J. & R.J. Hartin. 2000. PCR primers that
amplify fungal rRNA genes from environmental
samples. Applied and Environmental Microbiology
66: 4356-4360.
Redecker, D. 2000. Specific primers to identify arbus-
cular mycorrhizal fungi within colonized roots.
Mycorrhiza 10: 73-80.
Russell, A.J., M.I. Bidartondo & B. G. Butterfield. 2002.
The root nodules of the Podocarpaceae harbour
arbuscular mycorrhizal fungi. New Phytologist 156:
283-295.
White,T. J., T.D. Bruns, S.B. Lee & J.W. Taylor. 1990.
Analysis of phylogenetic relationships by amplifi-
cation and direct sequencing of ribosomal RNA
genes. pp. 315-322. In: M.A. Innis, D.H. Gelfand,
J.N. Sninsky & T.J. White (eds.) PCR protocols: a
guide to methods and applications. Academic Press,
New York.
