large enough to place them apart while the
sequence differences within species should be
small enough to bring them together in the same
group (Meyer and Paulay 2005).
The concept of DNA barcodes was originally
proposed for animals, and the mitochondrial
coxidase subunit I (COI) gene was accepted as an
authentic barcode to accurately identify animal
species (Hubert et al. 2003). But COI could not
be successfully applied in plants due to very little
variation in mitochondrial DNA sequences in
plants (Kress et al. 2005). This led to testing of a
number of non-mitochondrial genes and regions
in plants, i.e., nuclear and chloroplast DNA, as
potential candidates for reliable DNA barcodes.
Based on a number of different studies, CBOL
proposed seven plastid-markers including four
plastid-encoded genes (rpoB, rpoC1, rbcL, and
matK) and three noncoding spacers (atpF-atpH,
psbK-psbI, and trnH-psbA) to be used as barcodes in species identification of plants (CBOL
Group 2009).
Several studies have been published on the
identification and classification of duckweeds on
the basis of genotyping as nicely reviewed by
Appenroth et al. (2013). Attempts to use DNA in
barcoding and phylogenetic studies of Lemnaceae can be traced back to the last decade of
the previous century when Jordan et al.
(1996) compared three species of this family (L.
minor, L. valdiviana, and Lan. punctata) on the
basis of the rpL16 region. Other researchers
studied phylogenetic relations in different species
of duckweed using rpS16 gene intron sequences
(Martyrosian et al. 2009) and the trnL-trnF
intergenic spacer sequence (Rothwell et al.
2004). Few other studies have made such
attempts as reviewed comprehensively by
Appenroth et al. (2013). Then in a large-scale
study, Bog et al. (2013) sequenced the plastidic
rpS16 and rpL16 regions for 54 and 55 clones,
respectively, from the 11 species of the genus
Wolffia. The library of these barcodes, coupled
with the results from AFLP, could clearly identify some species including W. australiana, W.
columbiana, and W. brasiliensis. However, some
species, for example, W. globosa, were mixed
with W. borealis and W. neglecta.
The CBOL proposed plastid-markers’ barcodes for plant identification (plastid-encoded
genes (rpoB, rpoC1, rbcL, and matK), and three
noncoding spacers (atpF-atpH, psbK-psbI, and
trnH-psbA)) were tested for the first time by
Wang et al. (2010) for their potential use as
barcodes in species discrimination of the family
Lemnaceae (Wang et al. 2010). They used 97
accessions from 31 species representing all five
genera. The encoded genes, i.e., rpoB, rpoC1,
rbcL, and matK, were conserved and could not
prove good marker for this purpose, but the
noncoding spacers, i.e., atpF-atpH, psbK-psbI,
and trnH-psbA, showed more variability than the
coding genes. They concluded that the atpFatpH spacer can be used as a universal DNA
barcoding marker for species identification in
duckweeds. However, as pointed out by Borisjuk
et al. (2015), the study of Wang et al. (2010) was
not a complete barcode analysis for three main
reasons: (1) Only 31 duckweed species were
represented while six species were missing,
(2) only a single clone was used for several
species, and (3) sequence data of only 20 sampled species (out of 31 species) were used for
barcode analysis. Borisjuk et al. (2015) attempted to complete the survey work of Wang et al.
(2010) by including the missing data for the six
species and more clones to provide a more
complete database of the two barcodes (atpFatpH and psbK-psbI) for the family of duckweed.
By utilizing the sequences obtained in their own
study and over 300 sequences downloaded from
NCBI database for these two barcodes, they
constructed a barcodes library for the family
Lemnaceae that could successfully identify 30 of
the 37 known species of duckweed (Borisjuk
et al. 2015). Four species including L. valdiviana,
L. minuta, W. globosa, and W. lingulata could
not be identified by any of these two barcodes
(atpF-atpH and psbK-psbI), but the remaining
three species had some ambiguities (Borisjuk
et al. 2015).
The plastid DNA barcodes studied in the
identification of duckweeds so far give promising
results in inter-species identification but may not
be fruitful in intra-species discrimination of
ecotypes. For example, Feng et al. (2017)
5 Genetic Diversity and DNA Barcoding in the Duckweed Family
63
sequence differences within species should be
small enough to bring them together in the same
group (Meyer and Paulay 2005).
The concept of DNA barcodes was originally
proposed for animals, and the mitochondrial
coxidase subunit I (COI) gene was accepted as an
authentic barcode to accurately identify animal
species (Hubert et al. 2003). But COI could not
be successfully applied in plants due to very little
variation in mitochondrial DNA sequences in
plants (Kress et al. 2005). This led to testing of a
number of non-mitochondrial genes and regions
in plants, i.e., nuclear and chloroplast DNA, as
potential candidates for reliable DNA barcodes.
Based on a number of different studies, CBOL
proposed seven plastid-markers including four
plastid-encoded genes (rpoB, rpoC1, rbcL, and
matK) and three noncoding spacers (atpF-atpH,
psbK-psbI, and trnH-psbA) to be used as barcodes in species identification of plants (CBOL
Group 2009).
Several studies have been published on the
identification and classification of duckweeds on
the basis of genotyping as nicely reviewed by
Appenroth et al. (2013). Attempts to use DNA in
barcoding and phylogenetic studies of Lemnaceae can be traced back to the last decade of
the previous century when Jordan et al.
(1996) compared three species of this family (L.
minor, L. valdiviana, and Lan. punctata) on the
basis of the rpL16 region. Other researchers
studied phylogenetic relations in different species
of duckweed using rpS16 gene intron sequences
(Martyrosian et al. 2009) and the trnL-trnF
intergenic spacer sequence (Rothwell et al.
2004). Few other studies have made such
attempts as reviewed comprehensively by
Appenroth et al. (2013). Then in a large-scale
study, Bog et al. (2013) sequenced the plastidic
rpS16 and rpL16 regions for 54 and 55 clones,
respectively, from the 11 species of the genus
Wolffia. The library of these barcodes, coupled
with the results from AFLP, could clearly identify some species including W. australiana, W.
columbiana, and W. brasiliensis. However, some
species, for example, W. globosa, were mixed
with W. borealis and W. neglecta.
The CBOL proposed plastid-markers’ barcodes for plant identification (plastid-encoded
genes (rpoB, rpoC1, rbcL, and matK), and three
noncoding spacers (atpF-atpH, psbK-psbI, and
trnH-psbA)) were tested for the first time by
Wang et al. (2010) for their potential use as
barcodes in species discrimination of the family
Lemnaceae (Wang et al. 2010). They used 97
accessions from 31 species representing all five
genera. The encoded genes, i.e., rpoB, rpoC1,
rbcL, and matK, were conserved and could not
prove good marker for this purpose, but the
noncoding spacers, i.e., atpF-atpH, psbK-psbI,
and trnH-psbA, showed more variability than the
coding genes. They concluded that the atpFatpH spacer can be used as a universal DNA
barcoding marker for species identification in
duckweeds. However, as pointed out by Borisjuk
et al. (2015), the study of Wang et al. (2010) was
not a complete barcode analysis for three main
reasons: (1) Only 31 duckweed species were
represented while six species were missing,
(2) only a single clone was used for several
species, and (3) sequence data of only 20 sampled species (out of 31 species) were used for
barcode analysis. Borisjuk et al. (2015) attempted to complete the survey work of Wang et al.
(2010) by including the missing data for the six
species and more clones to provide a more
complete database of the two barcodes (atpFatpH and psbK-psbI) for the family of duckweed.
By utilizing the sequences obtained in their own
study and over 300 sequences downloaded from
NCBI database for these two barcodes, they
constructed a barcodes library for the family
Lemnaceae that could successfully identify 30 of
the 37 known species of duckweed (Borisjuk
et al. 2015). Four species including L. valdiviana,
L. minuta, W. globosa, and W. lingulata could
not be identified by any of these two barcodes
(atpF-atpH and psbK-psbI), but the remaining
three species had some ambiguities (Borisjuk
et al. 2015).
The plastid DNA barcodes studied in the
identification of duckweeds so far give promising
results in inter-species identification but may not
be fruitful in intra-species discrimination of
ecotypes. For example, Feng et al. (2017)
5 Genetic Diversity and DNA Barcoding in the Duckweed Family
63