Eva-12-187 2

Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

188 | GOUS et al.

1 | I NTRO D U C TI O N Taberlet, 2010), rpoC1 and trnH‐psbA (CBOL Plant Working Group,
2009). Another well‐utilized marker is the internal transcribed spacer
Our daily diet contains many plant products produced as a result of 2 (ITS2) region that is found between the 5.8S and 26S rRNA genes
pollination, such as fruits, vegetables, nuts and seed‐derived com‐ in plants (Chen et al., 2010; Yao et al., 2010). ITS2 has been used
modities. This crucial ecosystem service not only ensures food on as the DNA barcode in recent pollen barcoding studies (Bell et al.,
our tables, but also the diversification and maintenance of natural 2017; Keller et al., 2015; Richardson, Lin, Quijia, et al., 2015; Sickel
plant populations (Daily et al., 1997; Klein et al., 2007; Kremen et al., et al., 2015). ITS1 can be similarly useful at identifying plants to spe‐
2007). Studying the interaction between plants and their pollinators cies level (Wang et al., 2015). Generally, a multi‐locus approach to
has traditionally been done by field‐based observation (Johnson, identification yields better results due to increased discriminatory
1997; Wester, Stanway, & Pauw, 2009) and palynology (Dafni, 1992; power (CBOL Plant Working Group, 2009; Burgess et al., 2011 and
Wilcock & Neiland, 2002). These methods are tedious and time‐con‐ as reviewed in Bell et al., 2016).
suming, and require experts in the fields of palynology and taxon‐ Mixed origin, environmental samples, such as pollen, are charac‐
omy to identify both the pollen and the pollinator. Similar pollen terized by the presence of DNA from different organisms that may
morphologies, especially from closely related taxa, further compli‐ or may not be degraded. Next‐generation sequencing (NGS) tech‐
cate plant identification by palynology (Hargreaves, Johnson, & Nol, nologies now allow high‐throughput sequencing of complex DNA
2004; Mullins & Emberlin, 1997; Williams & Kremen, 2007). These libraries (Liu et al., 2012) including pollen (Bell et al., 2017; Keller
requirements have limited studies on plant–pollinator interactions et al., 2015; Kraaijeveld et al., 2015; Richardson, Lin, Quijia, et al.,
for many pollinator genera, particularly in species‐rich regions where 2015; Sickel et al., 2015). Coupling together NGS and DNA barcod‐
there is an abundance of both plants and pollinators. ing (metabarcoding) could improve on the identification assessments
Taxonomic activities in the areas of entomology and botany of pollen compared to traditional microscopic identification methods
drive pollinator and palynology‐related work, but studies are often employed by palynology.
conducted independently from each other. Samples are often col‐ In this study, we investigated the possibility of using a historical
lected for taxonomic purposes, such as species identification, dis‐ bee collection as a pollen source for ITS1, ITS2 and rbcL metabarcod‐
tribution pattern determination or identifying new introductions. ing and examined the usefulness of this approach to identify plant
Individual specimens are labelled with descriptive collection infor‐ species from small amounts of pollen carried by bee specimens col‐
mation, including collection date, location, collector and other rel‐ lected over 100 years ago.
evant information before being stored in collection (Pennisi, 2000).
Flower‐visiting animals housed within natural history collections
may have pollen on their bodies. Although flower visitors were likely 2 | M ATE R I A L S A N D M E TH O DS
not collected with the aim of utilizing the pollen that was inadver‐
tently collected along with the specimen, this pollen holds important
2.1 | Pollen sample collection from bee specimens
information on the plants visited by the insect visitor, the identity of
a possible pollinator and the plant community structure where the Selecting an appropriate bee species for this study was not only de‐
organism was collected. Additionally, a number of specimens from pendant on the availability of the species within the collection, but
the same area, but from different temporal points, can be used to also when specimens were collected. Pollen loads of the specimens
provide a chronological map of the area’s plant and pollinator history. can also vary depending on whether the individual bee was captured
Historic collections may therefore provide a meaningful resource to on their way to or on their way from a floral visit. Based on these
investigate, not only pollinator–plant interactions over time, but also criteria, Megachile venusta Smith (Megachilidae) specimens were se‐
plant community change over time, providing important information lected from the South African National Collection of Insects housed
on diversity and distribution. at Biosystematics, Plant Protection Research: Plant Health, of the
DNA barcoding allows for identification and classification of or‐ Agricultural Research Council (ARC), Pretoria, South Africa. The col‐
ganisms based on a short nucleotide sequence (Hebert, Cywinska, lection is utilized for taxonomic classification of indigenous bee spe‐
Ball, & deWaard, 2003). Ideal DNA barcodes have significant in‐ cies, and houses type specimens.
terspecific genetic variation and are flanked by conserved regions Megachile venusta is indigenous to southern Africa, and bee
for universal primer binding to allow easy amplification across a specimens used for pollen sampling were collected from different
wide range of taxa (Kress & Erickson, 2008). There is still debate biomes across South Africa over a period of 93 years (1914–2007).
on the optimal DNA barcode for plants (Dong et al., 2015; Ferri et Three bee specimens with visible pollen from each decade, starting
al., 2015; Kress, García‐Robledo, Uriarte, & Erickson, 2015), but the from the 1910s up to the 2000s, were selected for inclusion in this
ribulose‐1,5‐biphosphate carboxylase (rbcL) and maturase K (matK) study. No specimens with visible pollen were available for the 1930s
chloroplast genes have been suggested as good candidate genes to or 1950s, and these decades are thus not represented here. Only
target (CBOL Plant Working Group, 2009). Other chloroplast genes one specimen was available and included for the 1940s. Accession
and regions have also been used successfully to barcode plants and information of the bee specimens used in this study is provided in
pollen, including trnL (Kraaijeveld et al., 2015; Valentini, Miquel, & Supporting Information Table S1.

You might also like