We are searching data for your request:
Upon completion, a link will appear to access the found materials.
Some plants are said to be
self-fertile. An example is
Assuming that no cross-pollination happened with other plants, if a self-fertile plant such as
prunus tomentosaproduces a seedling, what DNA will the seedling have? Is the seedling's DNA an exact copy of the parent plant's DNA, or do the genes get rearranged?
Selfing (aka self-fertilizing) differs from cloning. When selfing occurs, the offspring is not an exact copy of the parent. When cloning occurs, the offspring is an exact copy (except for a few mutations) of the parent.
Selfing implies that an individual will produce two gametes (typically a spermatozoid and an ovule but that might be a bit more complicated) and these two gametes are fusing to give the zygote (egg or offspring if you prefer).
As a consequence, when selfing, meiosis is occurring (and therefore segregation and recombination) so that the offspring is not an exact clone of the parent but rather some kind of a rearrangement of the parent genome (with a few mutations of course).
Practical Work for Learning
Class practical or demonstration
You can extract DNA – to see what it is like – from some plant and some animal material using equipment and chemicals you might find in a kitchen. For more thorough analytical work, you need more control over the components of your chemicals, and it may be worth investing in a kit from one of the major suppliers. This is a rough and ready method that should give reasonable quantities of DNA from quite large quantities of material.
You can run this as a demonstration, or as small group work. Or you could prepare enough of each of several materials to allow groups to take samples from which they extract the DNA.
Apparatus and Chemicals
For each group of students:
Access to water bath at 60 °C (optional)
Test tube, 1, for each sample to be used
Ice cold ethanol (IDA), 10 cm 3 for each sample to be used
Wooden spill, straw, glass rod or inoculating loop, 1 per sample
For the class – set up by technician/ teacher:
Blender for each material to be used
or knives to chop material and a mortar and pestle to grind material for each working group
Ice bath to keep materials cool as necessary (Note 1)
Source/s of DNA (Note 2) – to produce 10-20 cm 3 of blended material per sample
Table salt, a pinch (or 1 cm 3 ) for every 300 cm 3 sample (Note 3)
Strainer for each material to be used
Detergent, 30 cm 3 for each 300 cm 3 of blended material to be processed
Protease, for example, pineapple juice, contact lens cleaner, pinch of meat tenderiser
Health & Safety
- Take care with ethanol IDA (see CLEAPSS Hazcard) – which is highly flammable and harmful through skin contact because of the presence of methanol.
- Protease (see CLEAPSS Hazcard) is harmful as a powder and irritant in solution. Wear eye protection and wash off skin promptly.
- Electrical equipment: Any electrical appliances used in the lab should be checked according to your employer’s systems. Use a blender dedicated for laboratory activities, not one that will be used later to process food for human consumption.
1 Using ice-cold ethanol and ice-cold water increases the yield of DNA. Low temperatures protect the DNA by slowing down the activity of enzymes that could break it apart. A cell’s DNA is usually protected from such enzymes (DNases) by the nuclear membrane which is disrupted by adding detergent. DNases in the cytoplasm would destroy the DNA of viruses entering the cell. Cold ethanol helps the DNA to precipitate more quickly. Chill the ethanol in a screwcap plastic bottle in the prep room freezer. Below 4 °C ethanol is below its flashpoint so this is safe even if your freezer is not spark proof.
2 You can use a variety of substances for this extraction. The original of this protocol recommended split peas, but onions, and fish eggs or fish sperm (milt) are commonly recommended. It is important to check that your source material contains enough DNA. Kiwi fruit, strawberries and bananas are often recommended, but it is reported (see NCBE article Discovering DNA in the Links section) that the white strands produced here are usually pectin rather than DNA. Kiwi fruit temptingly contain protease that could help to digest the proteins surrounding DNA and make the addition of further protease unnecessary. Some foods (such as grapes) contain a lot of water and will make a watery ‘soup’. In this case, go back to the first step and add less water. You need an opaque cell ‘soup’ for good yields. The amount of DNA you will get will depend on the ratio of DNA to cell volume rather than the number of chromosomes in your material. Plant seeds (such as peas) contain a high proportion of DNA.
3 The salt added helps the DNA to precipitate (as it clumps together) when it meets the ethanol phase.
4 It is important to allow time for each step to complete. The detergent must sit for at least 5 minutes to disrupt the cell membranes and nuclear membranes.
5 If you don’t think you can see any DNA, dip your stick or rod into the surface of the ‘soup’ and then move it gently upward into the ethanol layer. Also, look closely at the ethanol layer for bubbles – sometimes clumps of DNA are loosely attached to the bubbles. If you can leave the mixture for 30-60 minutes, you may see more DNA precipitate.
6 Confirm that what you have is DNA by using a stain for DNA. (It may well be a mixture of DNA and RNA.) Confirm that what you have is not pectin by adding pectinase. If it dissolves it was pectin!
SAFETY: Wear eye protection when handling the enzyme solution.
Avoid skin contact with ethanol and with enzyme solutions or powders.
Wash any spills off your skin promptly.
a Chill your ethanol by placing in a freezer for at least 2 hours, or overnight. Keep it on ice throughout the procedure (Note 1).
b Make a thick ‘soup’ by blending your source material with a little table salt and some cold water (Notes 1, 2, 3). For example, use 100 cm 3 of split peas, with 200 cm 3 of cold water and a pinch of table salt (around 1 cm 3 – Note 2). Blend on high for 15 seconds.
c Strain your ‘soup’ through a mesh strainer and collect the liquid part in a beaker.
d Add 2 tablespoons (30 cm 3 ) of washing-up liquid and swirl to mix.
e Let the mixture settle for 5-10 minutes (Note 4). Some protocols recommend carrying out this stage in a water bath at 60 °C. This may increase yield by increasing the breakdown of cell and nuclear membranes, or reduce yield if it stimulates the action of DNase enzymes (Note 1).
f Pour the mixture into test tubes or other small glass containers, to make each about one third full.
g Add some protease enzymes to each test tube. You could use a pinch of meat tenderizer, a few drops of fresh pineapple juice, or some contact lens cleaning solution.
h Tilt your test tube to 45° and slowly pour well-chilled ethanol (IDA) into the tube so that it forms a layer on top of your ‘soup juice’ – about the same volume as you have of ‘soup/ juice’. Ethanol is less dense than water and will float on top. DNA is soluble in water, but salted DNA does not dissolve in ethanol and will form white clumps where the water and ethanol layers meet (Note 5).
i Use a wooden stick or a straw (or a glass rod) to collect the DNA. Dip the stick into the tube and touch the white layer. Twirl the rod and the DNA should ‘spool’ onto the rod. DNA is a long, stringy molecule. As you pull on one end of the strand, it pulls more DNA into the ethanol layer where it will precipitate. (Note 3 and 5)
j Dry the sample on a paper towel if you want to measure the mass of product, or simply save the DNA by placing it in ethanol in any suitable small container with a lid.
Each cell in the human body contains 46 chromosomes. If you unravelled the DNA from each chromosome and put the 46 segments end-to-end, each cell would contain about 2 metres of DNA. Each piece of DNA is around 4-5 cm long.
What happens at each step?
- blending with salt and water: breaks the cells apart from one another and increases the surface area exposed to reagents such as detergent. It also begins to disrupt some of the cell walls of plant material.
- adding salt: means the DNA is more likely to clump together when it meets the ethanol layer.
- adding meat tenderiser: meat tenderiser commonly contains bromelain or papain – protease enzymes extracted from pineapple and papaya respectively. It will digest proteins associated with the DNA and so may help to purify the sample.
- which sources give the most DNA? how will you measure the sources and the product accurately for comparison?
- do you get better yields if you keep things cool throughout, or if you heat the blended mixture with detergent at 60 °C?
- which detergent works best?
- does the meat tenderiser make a difference? with every source?
- try extracting DNA from things which might not contain it
- how could you prove that this was indeed DNA? (what is the effect on it of stains which act on DNA such as toluidine blue or acetic orcein?)
Several companies produce kits for DNA extraction (for example Edvotek and NCBE – links below). The advantage of using kits is that any enzymes, salts, surfactants and buffer solutions provided will be pre-tested to ensure consistency from batch to batch, and hence reliability of the outcome of the procedure. They will also probably be cheaper than if you tried to source such materials directly. For a simple ‘bulk’ extraction like this, using many domestically available chemicals, it may not be necessary to use a kit. However, if you want clean DNA for further analysis, the kits are recommended. Some kits allow students to extract DNA from their own cheek cells and to encapsulate that DNA in a small plastic pendant (on a chain). Many students respond very positively to this personal dimension in the protocol, and it may make the procedure more memorable.
Health & Safety checked, March 2009
This website provides a clear, simple procedure and useful FAQs.
Edvotek is a biotechnology education company providing kits and workshops to support teaching and learning in biotechnology. They produce a simple DNA extraction kit (What does DNA look like?) and a kit (Genes in a tube) for extracting DNA from student cheek cells and making the sample of DNA (in a microcentrifuge tube) into a pendant.
The NCBE has an international reputation for developing innovative educational resources and making sophisticated biotechnology techniques accessible to classroom teachers and students. They provide courses and materials for a wide range of biotechnology practicals – including a DNA pendant kit using which students can store the DNA collected from their own cheek cells in a glass vial. Their website also contains a link to a protocol for extracting DNA from frozen peas (at www.ncbe.reading.ac.uk/NCBE/PROTOCOLS/unpublished.html).
The article Discovering DNA (in particular a paragraph headed DNA your onions?) on the NCBE site (www.ncbe.reading.ac.uk/DNA50/peadna.html) explains that fruits yield pectin rather than DNA.
(Websites accessed October 2011)
© 2019, Royal Society of Biology, 1 Naoroji Street, London WC1X 0GB Registered Charity No. 277981, Incorporated by Royal Charter
What’s the difference between plant DNA and animal DNA?
In the center of every plant cell, from algae to orchids – and in the center of every animal cell, from jellyfish to you and me – there’s a copy of the organism’s genetic material. This DNA carries a complete blueprint of the organism. It’s what transfers characteristics from one generation to the next.
There are pretty obvious differences between plants and animals, but – at the chemical level – the cells of all plants and all animals contain DNA in the same shape – the famous “double helix” that looks like a twisted ladder. What’s more, all DNA molecules – in both plants and animals – are made from the same four chemical building blocks – called nucleotides.
What is different is how these four nucleotides in DNA are arranged. It’s their sequence that determines which proteins will be made. The way the nucleotides are arranged, and the information they encode, decides whether the organism will produce scales or leaves – legs or a stalk.
Research shows that plants and animals may produce some proteins in common. One prominent example is known as Cytochrome C. But because the DNA copying process is imperfect, mistakes accumulate over time, making Cytochrome C slightly different in different creatures. The gene regions that specify the amino acid sequence in human Cytochrome C are more similar to those in another mammal like a rabbit, and less similar to a more evolutionarily distant creature, like a sunflower.
The schematic of classifying animals and plants in kingdoms is facing competition. More recently an alternative system has arisen, based on evolutionary and molecular information. Cytochrome c is perhaps the canonical or paradigmatic molecule in this approach.
Every species has a characteristic number of chromosomes, called the chromosome number. Animals have more chromosomes plants have fewer.
Tracking Genes for Self-pollination in Arabidopsis
Ultraviolet fluorescence microscopy image of a stigma of an arabidopsis plant that has been genetically modified to resist self-pollination. The left side was self-pollinated and the right side pollinated from another variety, and pollen tubes form only on the cross-pollinated side. Credit: Nasrallah Lab
Some plants need a partner to reproduce. Pollen from one plant pollinates the stigma of another, and a seed is formed. But other plants can self-pollinate, a handy survival mechanism for a lonely plant.
The ability to self-pollinate turns up in cultivated tomatoes and canola, among other important crops, and sometimes it can be a nuisance for plant breeders and seed producers who want to develop highly desirable hybrid varieties and produce hybrid seed on a commercial scale. To get hybrid seed, they plant two different varieties in the same field to allow them to cross-pollinate. But if one or both varieties can self-pollinate, workers must remove the pollen sacs (anthers) from the flowers by hand to prevent "selfing." This is so labor-intensive that it is usually only done in countries where labor is cheap.
Now Cornell researchers are zeroing in on genes that turn a plant's ability to self-pollinate on and off. Their work is described in the May 1 issue of the journal Current Biology and in the journal's online edition.
"The long-term goal is to understand how self-pollination is inhibited in self-incompatible plants, which are unable to self-pollinate because their stigmas can recognize and reject their own pollen. Then you could transfer this ability to any plant and use it to make hybrids," said June Nasrallah, the Barbara McClintock Professor of Plant Biology at Cornell.
Nasrallah's research group is working with Arabidopsis thaliana , a plant related to cabbage and mustard that is widely used in plant genetic research and whose genome has been sequenced. Previously, the group showed that two genes known as SCR and SRK are the key to self-incompatibility. SCR codes for a protein on the surface of pollen grains, and SRK codes for a receptor in the cell membranes of stigma cells. When these two proteins come from the same plant, the stigma rejects the pollen, and fertilization does not occur.
A. thaliana is highly self-fertile, but the Nasrallah group inserted SCR and SRK genes from another species, A. lyrata , which is self-incompatible, and created A. thaliana varieties that ranged from self-incompatible to "pseudo self-compatible," where a plant resists self-pollination for a while, but if it is not pollinated from another plant it will eventually accept its own pollen. In nature, pseudo self-compatibility is a best-of-two-worlds mating strategy, Nasrallah said, because it maintains the benefits of out-crossing while providing reproductive assurance when mates or pollinators are scarce.
In the latest research, Pei Liu, a postdoctoral researcher in Nasrallah's laboratory, and colleagues mapped the genomes of several varieties of transgenic A. thaliana in fine detail and isolated a gene known as PUB8 that seems to regulate the expression of SRK -- that is, whether or not it is turned on to manufacture its protein. The PUB8 gene shows some variation from one variety of A. thaliana to another, i.e., the DNA sequence contains a few different bases here and there. The degree to which self-incompatibility is turned on in the plant seems to correlate with these variations. PUB8-mediated pseudo self-compatibility might have been a transitional phase in the evolutionary switch from self-incompatibility to selfing in A. thaliana , Nasrallah speculates.
PUB8 is very close to SCR and SRK on the genome. It is unusual to find a regulatory gene so close to the gene it regulates, the researchers noted. PUB8 is expressed in other parts of the plant and probably has other functions, they said, adding that still other genes are probably involved in self-incompatibility.
Transformation of Arabidopsis with a Brassica SLG/SRK region and ARC1 gene is not sufficient to transfer the self-incompatibility phenotype
Self-incompatibility (SI) promotes outbreeding in flowering plants, and in Brassica SI is genetically controlled by the S locus. Self-incompatible Brassica and self-fertile Arabidopsis belong to the same crucifer family. In addition, a comparative analysis reveals a high degree of microsynteny between the B. campestris S locus and its homologous region in Arabidopsis--with the notable exception that the Brassica SI genes, SLG and SRK, are missing. Brassica ARC1 encodes a component of the SRK signal transduction pathway leading to self-pollen rejection, and no closely related ARC1 homolog has been identified in Arabidopsis. The purpose of the research reported here was to introduce Brassica SI components into Arabidopsis in an attempt to compensate for the missing genes and to investigate whether the SI phenotype can be transferred. Inserts of approximately 40 kb from the fosmid clones F20 and F22, which span the B. napus W1 SLG-SRK region, were cloned into the plant transformation vector pBIBAC2. Transgenic plants were generated that expressed the Brassica SI genes in the flower buds. In addition, the endogenous, SLG-like, gene AtS1 was not co-suppressed by the Brassica SLG transgene. No SI phenotype was observed among the T1 BIBAC2-F20 and BIBAC2-F22 transgenic plants. When the ARC1 gene was transformed into BIBAC2-F20 or BIBAC2-F22 plants, the resulting BIBAC2-F20-ARC1 and BIBAC2-F22-ARC1 plants still set seeds normally, and no rejection response was observed when self-incompatible B. napus W1 pollen was placed on BIBAC2-F20-ARC1 or BIBAC2-F22-ARC1 Arabidopsis stigmas. Taken together, our results suggest that complementing Arabidopsis genome with Brassica SLG, SRK and ARC1 genes is unlikely to be sufficient to transfer the SI phenotype.
Get it to Glow: Measuring Success
Since the process of bacterial glowing was discovered, scientists have tried to apply it to other organisms as well. One example of this is Green Fluorescence Proteins (GFPs). GFPs are used by scientists to tell if experiments involving gene exchanges have worked. The genes for GFPs are included when scientists insert other genes into an organism’s DNA. If the organism glows after these genes were inserted, then scientists know that the insertion was successful.
A glowing plant. This is what the successfully modified plants should look like in the dark.
In the same way, modifying a plant’s blueprint to turn on the production of luciferins and luciferase should allow for the plant to glow just like some bacteria do. Scientists tried to insert the DNA for the lux operon (from bacteria) into the DNA of the plants’ sex cells. They predicted that if the DNA for the lux operon was correctly inserted, some of the seeds from that plant would develop into plants that would glow.
When inserting DNA, sections that code for antibiotic resistance are included to help scientists test for the correct product. If scientists expose the modified plants to antibiotics and the plants survive, then they know that they took up the new genes correctly.
Present-day fingerprinting of plants
Method development and choice of markers
PCR-based multi-locus methods
Shortly after the invention of the ingenious PCR procedure by Saiki and colleagues , three PCR-based approaches to generate DNA fingerprints were published more or less at the same time. All of these methods used single oligonucleotide primers with arbitrary sequences to produce PCR fragments from genomic DNA, resulting in multi-locus banding patterns after electrophoretic separation and visualization by staining or radiography [50-52]. The so-called random amplified polymorphic DNA (RAPD) approach developed by Williams and colleagues  soon became the most popular variant of these methods. Major explanations for this immediate success include the small quantities needed of sample DNA, and the simple and fast procedures compared to the hybridization-based methods. Results from a typical RAPD experiment are illustrated in Figure 2 .
Random amplified polymorphic DNA (RAPD) fingerprints of Pelargonium. DNA aliquots from 13 cultivars of Pelargonium hortorum (lanes a to m), nine cultivars of P. peltatum (lanes n to v), and one individual of the wild subspecies P. peltatum ssp. dibrachya (lane w) were amplified with the arbitrary 10-mer primer OPG-5. RAPD products were separated on a 1.5% agarose gel and stained with ethidium bromide. Positions of size markers (lane S) are indicated in base bairs (bp). N, negative control (no template DNA in the PCR assay). MW, molecular weight.
A few years later, Zabeau and Vos  and Vos and colleagues  presented the amplified fragment length polymorphism (AFLP) technique, which represented an ingenious combination of RFLP and PCR methodology. AFLP analyses became soon very popular, mainly because of the large numbers of polymorphic bands obtained in a single experiment. The inter-simple sequence repeat (ISSR) method developed by Gupta and colleagues  and Zietkiewicz and colleagues  relied on microsatellite-complementary PCR primers that could be used in an anchored or unanchored version. RAPD, AFLP and ISSR are still much used nowadays, although RAPD especially has often been criticized for problems with reproducibility and competitive priming, as reviewed in Weising and colleagues . These problems are less pronounced for AFLP and ISSR where more stringent PCR conditions can be applied. Nevertheless, all three methods usually arrived at quite similar estimates of genetic diversity and genetic distances when applied to the same plant material, as reviewed in Weising and colleagues .
Other less frequently used methods to generate multi-locus PCR fingerprints include the sequence-related amplified polymorphism (SRAP) technique that specifically amplifies polymorphic junction fragments between exons and the flanking intronic DNA [57,58], and the target region amplification polymorphism (TRAP) method . Common features of SRAP and TRAP include the use of two primers of about 18 nucleotides length (one of which targets a protein-coding region), and non-stringent PCR conditions during the first five cycles. The so-called selective amplification of polymorphic microsatellite loci (SAMPL) is a variant of the AFLP technology that combines AFLP- and microsatellite-specific primers , whereas the direct amplification of minisatellite DNA (DAMD) utilizes primers that are specific for minisatellites rather than microsatellites . Yet another approach, resistance gene-analog polymorphism (RGAP), makes use of PCR primers that bind to the conserved domains of plant resistance genes .
The Diversity Arrays Technology (DArT) is a high-throughput method based on the hybridization of fluorescent DNA probes to a set of target DNAs spotted onto a microarray [63,64]. The DNA is first digested with one or two restriction enzymes, followed by the ligation of specific adapters as in AFLP. Individual PCR products are spotted onto a grid to form an ordered microarray that represents hundreds of arbitrarily selected restriction fragments from all cultivars/species and various genomic regions of the gene pool of interest. Individual genomic DNA samples are pretreated in the same way as the pooled representatives (that is, restriction, ligation of adapter, and PCR with adapter-specific primers). Before being individually hybridized to the chip, each probe DNA is labelled with a fluorochrome to enable detection. Like AFLP and RAPD, DArT does not require previous sequence information. It allows simultaneous analysis of hundreds or even thousands of polymorphic loci, but the need to generate a microarray restricts the general use of the technique. By 2012, DArT technology has been developed for about 60 organisms, mostly crop and model plants , but also some wild plants such as the fern Asplenium viride.
Transposable elements and especially the retrotransposons bounded by long terminal repeats (LTRs) have proved to be useful for developing particularly sensitive multi-locus profiling techniques, either alone or in combination with other types of primers . In the inter-retrotransposon amplified polymorphism (IRAP) approach developed by Kalendar and colleagues , primers are directed towards the LTRs of BARE-1, a retrotransposon of barley. The same authors also introduced retrotransposon-microsatellite amplified polymorphism (REMAP) which combines outward-facing LTR-specific primers with anchored microsatellite primers. Basically the same strategy, known as copia-SSR, was simultaneously developed by Provan and colleagues . In the so-called sequence-specific amplification polymorphism (S-SAP) analysis, retrotransponson-specific primers are combined with AFLP primers . S-SAP often produces highly variable fingerprints that are frequently more informative than AFLP. Related approaches have been developed for other plant transposons [71,72].
PCR-based single-locus methods
Because of their abundance, high polymorphism in the number of tandem repeats, co-dominant inheritance, excellent reproducibility and ease of use, PCR-amplified single-locus microsatellite markers have become the marker of choice for many applications, and presently remain more important than any of the other traditional DNA fingerprinting methods [73,74]. Typically, a pair of microsatellite-flanking primers is used to amplify the targeted locus by PCR, amplification products are separated by polyacrylamide or capillary electrophoresis, and banding patterns are monitored by radiography or fluorography. When locus-specific microsatellite analysis was first used in plants in 1992 , the need for developing species-specific microsatellite-flanking primers was still a serious drawback, requiring tedious cloning and enrichment strategies (see the reviews by Squirrell and colleagues  and Weising and colleagues ). Nowadays, this task has become relatively simple for (1) the increasing number of plant species with DNA sequence data in public databases and (2) the development of ultrafast “next generation sequencing” technologies that enable the identification of microsatellite loci and design of primers by random genomic sequencing (see "The future of DNA fingerprinting" below). A typical result from a microsatellite genotyping experiment is shown in Figure 3 .
Locus-specific microsatellite analysis of four populations of the Lesser Periwinkle (Vinca minor) using a primer pair specific for locus Vimi43 (Möller, personal communication). For each population, ten samples were genotyped. Populations 1 and 2 were collected in northern Italy, whereas populations 3 and 4 were sampled in central Germany. Strong indications for clonality can be found in populations 1, 3 and 4. S, size standard: T-ladder derived from a chloroplast DNA fragment of Macaranga indistincta. Molecular weights (MW) of size markers are indicated in base pairs (bp).
Lately, expressed sequence tags (ESTs) have become a viable alternative to genomic DNA as a source for SSR loci, resulting in so-called EST-SSR markers that are either generated by cDNA cloning and sequencing  or, more commonly, by exploiting existing EST databases [78-80]. Database mining is often quite efficient, since EST-SSRs are surprisingly common and may be expected every 2 to 10 kb of EST sequence for example, one per 6.3 kb in Hordeum vulgare. These estimates of course depend on the search criteria and the search script used, most importantly on the minimum number of repeats used to define a microsatellite. Trinucleotide repeats commonly prevail in protein-coding regions of ESTs, whereas dinucleotide repeats are more frequent in 5′ and 3′ untranslated regions (UTRs). Expansions and deletions in coding regions can be tolerated for tri- and hexanucleotide repeats, because they do not perturb reading frames.
EST- and cDNA-derived SSRs have several important advantages over anonymous markers (see the review by Varshney and colleagues ). First, developing markers from already existing sequences is easy, fast and economical. Second, any type of microsatellite will be detected, whereas only SSRs with predefined motifs are captured by enrichment strategies. Third, EST-SSRs are physically linked to an expressed gene, which may encode a trait of interest. Finally, primer target sequences that reside in transcribed DNA regions are expected to be relatively conserved thus enhancing the chance of marker transferability across taxa. On the negative side, the association with coding regions sometimes limits the polymorphism of EST-derived SSR markers, resulting in fewer alleles and/or lower observed heterozygosity , but this is not necessarily the case. For example, Pashley and colleagues  compared the performance of 48 anonymous versus 48 EST-derived SSR markers from common sunflower, Helianthus annuus, and their transferability to two other Helianthus species. Their study showed that: (1) 73% of the EST-derived SSR markers were transferable among all species, compared with only 21% of the anonymous SSR markers (2) EST-SSRs were on average only slightly less polymorphic that anonymous SSRs, both in the focal and the non-focal species and (3) EST-SSRs located in coding regions were more readily transferable than those in untranslated regions - without differing significantly from the latter in terms of variability.
Locus-specific markers can also be developed from individual bands within multi-locus profiles, as exemplified by the sequence characterized amplified regions (SCARs). In the original description of the approach, specific primer pairs were designed for amplifying single bands of a RAPD profile . SCARs have been used for cultivar identification in, for example, olive, Olea europaea, and sweet cherry, Prunus avium. In the cleaved amplified polymorphic sequences (CAPS) approach , the resulting PCR product is treated with a restriction enzyme before scoring of fragments.
Single nucleotide polymorphisms
In the last decade, DNA fingerprinting methods based on single nucleotide polymorphisms (SNPs) have become increasingly important, especially in conjunction with microarray analyses that allow the simultaneous screening of very large numbers of SNP sites (see the review by Appleby and colleagues ). Among the many types of mutations occurring in genomes, single nucleotide exchanges stand out by their high absolute numbers as well as their biallelic nature, relatively low mutation rates, even distribution across the genome and relative ease of detection. In plants, one SNP is typically found per approximately 100 to 500p of DNA, but the average density depends on the studied species and the genomic region investigated. Numerous technologies have been developed for SNP discovery as well as for SNP genotyping . Direct sequencing of multiple copies of the same genomic region is the most obvious method for SNP discovery, and has become very efficient after the development of high-throughput sequencing systems [87,88] (see "The future of DNA fingerprinting" below). Like SSRs, SNPs can also be mined from existing databases . Practically all SNP genotyping assays are amenable to automation and therefore allow routine high-throughput analyses of large numbers of samples.
SNP markers are already well established in all major crop species [90,91], especially in those for which fully sequenced genomes are available. Recently, microarrays with typically 10,000 to 40,000 SNP markers (SNP-Chips) have been developed for many crops, and large-scale screenings of germplasm collections can now be undertaken at comparatively low costs . Given that thousands of SNPs can be detected by novel sequencing approaches, SNP genotyping will receive increased attention, at least in economically important crop plants. However, poor transferability to related species may hamper their successful implementation for large-scale genotyping projects across genera.
Organellar DNA-based methods
The most commonly used organelle for genetic studies in plants is the chloroplast. Since recombination is rare or absent in plastid genomes, all DNA polymorphisms for a certain individual can be combined to form a “haplotype”. Plastid DNA polymorphisms at the intra-specific level are relatively rare, and the numbers of detected band profiles (haplotypes) are therefore often considerably lower than those detected by nuclear markers. On the positive side, the high conservation of organellar DNA sequences has enabled the development of non-specific, so called universal, PCR primers that amplify cpDNA introns and intergenic spacers in a wide array of plant species [93,94]. Universal primers are also available for the amplification of SSR loci in the chloroplast genome . Polymorphisms within the amplified fragments can be monitored by various approaches, including the detection of length variants by high-resolution electrophoresis, and the detection of sequence variants by sequencing, or by digesting the PCR products with restriction enzymes in an approach called PCR-RFLP.
Plastid DNA is especially useful in studies where a low mutation rate is desirable, such as in the analyses of phylogenetic and phylogeographic patterns. Often both plastid and nuclear markers are combined in the same study for complementary information. Since the mode of plastid inheritance is usually maternal in angiosperms and paternal in gymnosperms, these markers also have the potential for tracing uni-parental lineages over large distances in time and space.
Choice of method
The pros and cons of different molecular marker methods have been discussed in a number of comparative investigations (see [27,96]). The actual choice of method must of course take marker availability, costs, expertise, equipment and many other factors into consideration. Based on 292 papers published between mid-2006 and mid-2009 on discrimination among plant cultivars, locus-specific microsatellite analysis (SSR) was the most popular method (36%), followed by RAPD (27%), ISSR (13%), AFLP (11%), other nuclear DNA-based methods (10%, including for example CAPS, DAMD, IRAP, REMAP, SNPs, SCAR and SRAP) and organellar DNA-based methods (3%, mostly cpDNA) . If the purpose of a study is to simultaneously discriminate both dissimilar and very similar entities, applying a whole battery of marker types may be the best solution.
While insufficient repeatability of DNA marker profiles can be regarded as a methodological artefact, insufficient germline stability of sequences corresponding to DNA markers can cause 𠇋iological artefacts” due to excessively high mutation rates. This problem is most likely to arise with the most sensitive types of markers, such as SSRs. The ability to merge data from different studies, even when developed in different laboratories, is a major asset of this method. The same is true for the other single-locus DNA markers, such as SNPs, SCARs and CAPS, but these are usually only biallelic. Nevertheless, the potential number of SNPs is virtually unlimited, and various SNP-based assay methods have already been developed (see above). In a comparative study on 58 maize inbred lines, SNPs outperformed SSRs both in terms of quality and quantity .
Exceptionally high mutation rates and reduced germline stability are often encountered when using retrotransposon-based markers . Thus, several reports have indicated that S-SAP markers are especially useful for discriminating among clones derived by somatic mutations  or among genotypes derived by recombination among highly similar entities . The S-SAP primers are usually designed according to species-specific sequence information but positive results have also been obtained by using universal retrotransposon-based sequences .
Besides their application for the identification of plant material and for the estimation of similarity and relatedness, DNA markers have been extensively used for the construction of genetic linkage-based genomic maps, with a major aim of identifying markers that are closely linked and therefore co-inherited with genes for specific traits (see "Genetic mapping" below). Dense linkage maps have been constructed for numerous plant species including all major crops using all kinds of markers. For ease of scoring when screening large numbers of progeny, singe-locus biallelic markers such as SNPs are usually preferred for this purpose.
Applications of present-day DNA fingerprinting in plants
Since the humble beginnings in 1988, DNA fingerprinting has become an immensely important instrument for genotype identification in both wild plant species and their cultivated relatives. Plants differ widely in life history traits including reproductive parameters such as propagation method and, for those that propagate by seed, also in breeding system (selfing or cross-pollinated) and in the mode of pollen and seed dispersal. All these factors have profound influence on the amount and partitioning of genetic variability between and within various entities such as cultivars and populations. These differences affect the utilization of DNA markers for fingerprinting individual plants or genotypes.
In some cases, DNA-based estimates of similarity among a set of genotypes show a relatively close association with previous morphology-based estimates, but there are also considerable discrepancies in other cases. If the morphological characters are mostly quantitative in nature, correspondence with DNA marker estimates is generally quite high as compared to qualitative characters, which are more likely to reflect only a small number of mutation events. It has also been suggested that molecular data are better at differentiating cultivated genotypes as well as their wild relatives according to origin and pedigree, whereas conventional pomological characterization data are more closely associated with physiological properties .
Genotype identification in wild plants: the influence of life history traits
Proper identification of individual genotypes is an important basis for many wild-plant-based studies. As mentioned above, various life history traits affect the amount and partitioning of genetic variation. Inbreeding species are, for example, most useful for forensic applications, since they typically produce suitable-sized patches of genetically identical or almost identical plants . By contrast, outcrossing species are characterized by a situation where every plant has a different genotype. While potentially very informative, it is usually extremely difficult to secure forensic evidence involving a particular, unique plant specimen. Clonal plants, whether due to extensive vegetative propagation or apomixis frequently produce large numbers of progeny with the same genotype (see also Figure 3 ). Such genotypes can cover large geographical areas and are thus not sufficiently accurate for tying botanical evidence to a certain location.
In other research areas, the variation in plant life history traits can, however, be regarded as a positive factor a wide range of biological questions can be answered by choosing suitable material and methods. DNA marker analyses have thus been able to estimate genotype age in plant clones, which have often proved to be considerably larger - and therefore often also older - than expected from previous data. For example, Steinger and colleagues  studied Carex curvula, a sedge species found in the European Alps. RAPD analysis of 116 tillers from a small patch (2.0 ×𠂐.4 m) identified a total of 15 multi-locus genotypes. More than half of the sampled tillers proved to belong to a single, large clone estimated to be around 2,000 years old. Invasive species sometimes produce particularly large clones, such as the Japanese knotweed, Fallopia japonica, and the alligator weed, Alternanthera philoxeroides, both of which displayed a single RAPD phenotype in spite of being sampled over very large areas [105,106]. In other cases, DNA marker analyses have revealed more heterogeneity than expected. Each of five investigated Chinese populations of the invasive water hyacinth Eichhornia crassipes were thus shown to consist of at least three different clones according to their RAPD profiles .
Information about clonal growth can be very helpful for determination of factors involved in shaping population structure. When a microsatellite-based study was carried out in the marine eelgrass Zostera marina, clonal size proved to be positively correlated with heterozygosity . Outbreeding clones were larger and contained more flowering shoots, indicating that inbreeding depression had decreased vigor and fertility. An unexpectedly high degree of genetic homogeneity was recently described in the geophyte Gagea spathacea. All but two of 138 examined specimens, representing 52 populations throughout the entire distributional area in northern, central and eastern Europe, had identical AFLP profiles. Probably this highly polyploid taxon has derived from a hybridogenic event, and has managed to attain its large area almost exclusively by bulbil production and spreading rather than by seed set and seedling establishment.
DNA fingerprinting has also helped to clarify the reproductive system in species that can produce seeds both sexually and asexually (that is, by apomixis). Many Taraxacum populations consist of triploid individuals that apparently reproduce through apomixis and therefore are clonal. Such clones can sometimes cover large areas as was demonstrated in an AFLP study . A comparison of SSR and AFLP data showed that both marker types were able to discriminate among nine apomictic microspecies (defined on morphological characteristics) of Taraxacum, but that AFLP was more sensitive in detecting also small, mutation-derived differences within each microspecies . By contrast, two dinucleotide repeat SSR loci detected considerably more variation than AFLP in apomictic lineages of Ranunculus carpaticola[112,113]. Evidence for an origin by mutation instead of by recombination was provided by the lack of allele segregation in the investigated SSR loci. Thus, in each lineage, the same number of alleles was always found within a locus, and these alleles also formed classes of related allele sizes within each lineage.
The availability of adequate tools to identify individual genotypes can be immensely useful in plant ecology. Thus, SSR-analysed Taraxacum clones were recently employed to investigate biodiversity and ecosystem functioning. In one study, five identified Taraxacum clones were used for setting up experimental plots where effects of levels of diversity in both favorable (fallow field) and unfavorable (mowed lawn) conditions could be quantified . The genotypic diversity effects appear to be stronger in environments where intra-specific competition is more intense. In a parallel study, genotype ×𠂞nvironment associations were studied in natural populations with the same set of Taraxacum clones . Genotypes that produced poorly under favorable conditions instead showed the highest performance under stressful conditions.
Genotype identification in vegetatively propagated cultivars
All plants belonging to a particular cultivar of an asexually propagated crop are expected to share identical DNA fingerprints, except for rare mutations. By contrast, sexually derived cultivars are expected to exhibit non-uniform fingerprint patterns. However, there are crops for which the major breeding method involves selection amongst rather similar seedlings that originate from a very small number of widespread cultivars. This situation can be exemplified by peach, Prunus persica, which is self-fertile and self-pollinating to a large extent. In such crops, new cultivars sometimes have DNA fingerprints that are almost identical or at least very similar to those of the seed parent in spite of being derived through sexual recombination. By contrast, variation is sometimes encountered where one expected uniformity. Vegetatively propagated crops are usually still capable of producing sexually derived seeds, and these may germinate and develop into fertile but unnoticed plants in less well-tended fields and orchards. There is therefore an increased risk, especially for older cultivars, that a certain name is being used on several different entities, some of which have originated from seed setting.
Irrespective of propagation and breeding method, the value of accessions in plant genetic-material collections benefits tremendously from DNA marker-aided identification. This is, however, especially important in vegetatively propagated crops that must be grown in the field or maintained in greenhouses at high costs. Previous reviews [97,116] show that a higher number of mislabelled plant accessions are revealed using DNA markers (typically 25 to 30% mislabellings) as compared to traditional (pomological or ampelographic) characters (typically 5 to 10% mislabellings). Different categories of problems with synonyms and homonyms have been defined , and appear to be especially prevalent in locally grown and often older germplasm whereas well-known modern-day cultivars are correctly identified to a much higher extent.
For large-scale profiling of, for example, accessions in a genetic resources collection, SSR markers are usually preferred . Although generally regarded as highly reproducible, problems are sometimes encountered with incorrect allele sizing, the occurrence of null alleles, allele drop-out (only one of two alleles is amplified at a heterozygous locus), false alleles (artefactual amplification products) and occasional amplification of isoloci (an isolocus is a similar but non-identical locus in the genome, common in allopolyploid species). While dinculeotide repeat SSRs are the most common type of microsatellite, less stuttering and improved allele sizing can be achieved with markers based on tri- or tetranucleotide repeats [117,118], although these are sometimes also less informative.
Zhang and colleagues  checked the accuracy and reliability of 15 SSR loci for clone identification in cacao, Theobroma cacao, and reported an average error rate of only 0.014 for allele drop out and 0.019 for false alleles. Some loci were more error-prone than others, suggesting that putative loci should be evaluated not only for their polymorphism but also for reliability prior to large-scale analyses. Vélez and Ibánez  checked 19 SSR loci in a study of more than 4,000 plants representing 19 grapevine cultivars. After removal of some minor technical artefacts, 99.8% of the samples matched to the expected genotype. Some loci, however, proved to be rather sensitive to the occurrence of chimeric mutations whereas others were not. Artefactual variation of SSR markers was also indicated in a study of olive . Interestingly, SSR alleles that differed among olive samples from the same cultivar were only 2p (one repeat unit) apart, whereas samples from different cultivars usually exhibited larger size dissimilarity in the polymorphic alleles. If available, accurate pedigree information is very valuable for checking the reliability of marker profiles.
The possibility to merge SSR-derived data from different investigations is often reported as a major asset of this method. This option is, however, dependent on the use of identical SSR loci and suitable standardization procedures. Since absolute allele sizes of the SSR markers often differ when results from different laboratories are compared, a representative reference material with many different alleles should be used at all laboratories involved in the genotyping program, and the material for these standards should be harvested from predetermined plants in one collection only. By comparison with suitable standard alleles, sample alleles can then be defined according to relative number of core repeat units instead of relying only on the absolute fragment length in base pairs.
Increased attention has recently been paid to the use of SNP markers for genotype identification in vegetatively propagated cultivars. Advantages of SNPs are their potential abundance and the fact that they do not rely on fragment length variation like SSR, and therefore are easier to standardize across different laboratories and equipment. Numerous high- and low-density SNP arrays have recently been developed for different crops. For example, a set of 48 SNPs was developed in grapevine, Vitis vinifera, through resequencing of 11 genotypes . High-throughput SNP genotyping can be conducted using bead arrays or microarrays (SNP chips) such as in, for example, Citrus. Since the number of polymorphisms covered in these assays is usually several hundred to many thousand, the obtained data can also be used for detecting quantitative trait loci (QTL).
Genotype identification in seed-propagated cultivars
In seed-propagated crops, at least some genetic variation usually persists also within cultivars. This is especially pronounced in highly outcrossing species, thus making DNA-marker-aided cultivar identification considerably more difficult. The situation is further complicated by the fact that each seed production cycle can lead to the introduction of new genetic variation - for example, due to foreign pollen. A considerable influx of new alleles was thus demonstrated after 7 to 13 subsequent regenerations of open-pollinating rye, Secale cereale.
Even with all the precautions taken in connection with modern gene bank regenerations, changes in allele frequencies can result from just recombination and selection. This was clearly demonstrated in an AFLP analysis of 50 white cabbage, Brassica oleracea, accessions together with first-generation regeneration products from six of these accessions . The genetic changes between original accessions and their respective regenerants were of the same magnitude as the differences among some of the more similar accessions. Moreover, while most alleles remained stable between generations, frequencies of some alleles instead changed considerably, suggesting that unintentional selection had taken place.
Obviously, a large number of markers are required for proper quantification of genetic changes between generations, and for efficient discrimination among outcrossing, seed-propagated cultivars. Such large numbers are, for example, provided by the DArT technology that proved very useful for distinguishing Festulolium cultivars (Festuca × Lolium experimentally produced hybrids) with 7,680 probes on a microarray . In this study, each cultivar was represented by 20 individual plants. These plants were analyzed both as individual and bulked samples. In order to minimize the loss of low-frequency bands, bulks with only five plants in each were recommended.
Inbreeding crops are usually considered less problematic than outcrossers, since the cultivars are more homogeneous. However, some inbreeding crops still contain intra-varietal variability, especially in the case of primitive cultivars or landraces. Propagation cycles performed in a genebank with such material can cause prominent gene frequency changes due to gene flow and inadvertent selection. In these cases, pure-lining of the accessions may be necessary to avoid loss of diversity, as exemplified by the USDA Soybean Germplasm Collection . In addition, selfing crops often contain a multitude of genetically very similar cultivars, thus necessitating the use of highly polymorphic markers for discrimination. While the commonly applied AFLP and SSR markers have produced sufficient results in many studies, the retrotransposon-based S-SAP method has been shown to resolve even very closely related plant accessions in, for example, wheat, Triticum aestivum. Nowadays, SNP markers receive increasing attention also in sexually propagated crops, mainly for the almost inexhaustible number of potential polymorphisms. Genome-specific SNPs have thus been developed from wheat gene intronic regions, and have proven highly useful for cultivar discrimination as well as enabling a quantification of genetic diversity at each of the genomes in this hexaploid crop .
Genotyping somatic mutations
Spontaneously occurring somatic mutations can give rise to so-called 'sports’. These deviate from the original cultivars in minor but economically important traits such as fruit color in fruit and berry crops, and flower or leaf color in ornamentals. Sports are difficult to distinguish with DNA fingerprinting since the markers usually cover only a minute part of the genome. In addition, chimeras are quite common - that is, mutations that occur in only one of the three meristematic cell layers in the apical meristem that differentiate into the various plant tissues. The existence of chimerism was very elegantly demonstrated in grapevine, Vitis vinifera, by Franks and colleagues . Although grapevine is a diploid species, some SSR loci occasionally showed three alleles when different sports were analyzed. It turned out that plants regenerated from cell layers L1 and L2, respectively, had different SSR alleles as well as different phenotypic characteristics. SSR analysis was used to identify chimeric clones also in 'Cabernet Sauvignon’ , 'Grüner Veltliner’  and 'Moscatel Galego Branco’  while clones of 'Pinot’ were successfully distinguished with the S-SAP method . In this study, three different retrotransposon-based primer pairs produced a total of 1,274 bands, one third of which were polymorphic and able to discriminate among all the 19 investigated clones.
S-SAP analysis has been successful for the genotyping of sports also in other crops, such as apple. Using 15 S-SAP primer combinations, five sports of 'Gala’ and one of 'Braeburn’ could be discriminated, both from each other and from the two original genotypes , whereas 24 SSR primer pairs generating a total of 64 alleles, and 35 AFLP primer combinations generating more than 1,000 bands, failed to do so. Based on two Ty1-copia LTR retrotransposons, a set of 19 bud sports of the apple cultivar 'Fuji’ were investigated with S-SAP . All sports obtained unique DNA profiles. Other retrotransposon-based methods can also be quite useful. In 24 sports of clementine, Citrus reticulata, application of eight IRAP primers produced a total of five polymorphic bands whereas RAPD (26 primers), ISSR (16 primers), AFLP (8 primer combinations), S-SAP (9 primer combinations) and SSR (9 primer pairs) revealed, at the most, one (S-SAP) or two (RAPD) polymorphisms [135,136].
Some studies found surprisingly high levels of marker polymorphism within cultivars, such as in olive, where clones have been selected and subsequently multiplied by vegetative propagation for centuries. In one study, 27 putative clones of 'Verdeal-Transmontana’ could be differentiated with ISSR  while even higher levels of polymorphism were encountered with RAPD (50% polymorphic bands) and ISSR (54%) in the screening of 120 putative clones of 'Cobrançosa’ . Possible explanations for these observations include a polyclonal origin, accumulation of somatic mutations over the long life-span of this woody species, and unnoticed establishment of sexual progeny in the orchards.
Genotyping in vitro-propagated material
Heritable somaclonal variation - that is, variation among regenerants due to somatic mutations - can be significantly enhanced by some micropropagation techniques. Although often regarded as an undesirable side-effect, these mutations can be valuable in crops that lack sexual reproduction (such as, for example, banana) or have very long generation cycles (such as, for example, palm trees). In general, axillary branches yield the most stable regenerants, followed by somatic embryogenesis and finally organogenesis. It is, however, impossible to predict whether markers will be able to find any variation in regenerated material, or what methods will prove to be most efficient.
Very few polymorphisms have generally been found in tissue culture regenerants. The extent of DNA marker polymorphism can, however, vary considerably between plant materials - even of the same species - as was shown by comparing the very uniform regenerants of the banana cultivar 'Prata Ana’  with the highly variable regenerants of cultivar 'Valery’ . When AFLP analysis was applied to regenerants of Helichrysum italicum, plantlets derived directly from leaves showed the same level of variability as plantlets that had passed through a callus stage . Although only 6.2% of a total of 449 bands were polymorphic, almost all plantlets differed from the original genotype in at least one band. The same band polymorphism was encountered in several plantlets in some cases, suggesting a hot spot of DNA instability. In another study, plant material of date palm derived from asexual embryogenesis showed considerably more variability than plants derived from organogenesis when analyzed with AFLP markers .
Detailed sequence-based analysis of the molecular events responsible for SCAR marker polymorphism (for example, insertion or excision of transposons, microdeletion, recombination) between somaclones and sexual recombination-derived lines of maize, demonstrated that the same mechanisms apparently determine both in vitro and in vivo variability . Therefore, it was concluded that cell culture only enhances the rate of heritable genomic changes which otherwise occur naturally in living organisms. Carrier and colleagues  studied somaclonal variation in the grapevine cultivar 'Pinot noir’ by high throughput sequencing and found that insertion polymorphism generated by transposable elements was responsible for most of the variation.
In theory, DNA fingerprints obtained from plant fragments should be able to provide important evidence in crime investigations but success has been limited so far, probably due to problems with isolating DNA of sufficient quality from poorly preserved plant material. SSR markers are often chosen for forensic work since they work comparably well also with heavily degraded DNA. One famous early case, however, involved RAPD analysis of seed pods of the Palo Verde tree, Cercidium sp., recovered both from the crime site and from the pick-up truck of a suspect , while another case made use of SSR and RAPD analysis to compare fragments from clonally reproducing bryophytes (mosses) collected both on the crime site and on the suspect himself . In subsequent experiments, a high likelihood of picking up fragments of bryophytes by walking outdoors wearing rubber boots was shown, as well as the ability to isolate DNA of sufficient quality after several months of storing bryophyte material under adverse conditions . These facts together with the high level of clonality in many bryophyte species make them an ideal target for forensic analysis. In yet another criminal case, seedlings of the inbreeding herbaceous knotweed Polygonum aviculare obtained from germinating seeds found in the wheelhouse of a suspect’s car tire, and from a large number of soil samples taken at the crime site and various reference localities, were analyzed with AFLP .
Detection of adulterations of food, drink and medicinal products is another area for forensic botany. Licensing arrangements sometimes require that a specified clone, cultivar or landrace is utilized in the manufacturing of food and beverages. Thus, well-defined grapevine clones must be used to receive 𠇊ppellation d'origine controllພ” labelling in France. In one study, musts (that is, freshly pressed grape juice destined for wine-making) from two different grape cultivars could be identified using two SSR markers . In another study, musts containing different proportions of two grape cultivars were analyzed with densitometry measurements of the SSR amplification products after separation and staining on polyacrylamide gels . In Greece, Nemea wines are marketed with protected denomination of origin (PDO). Instead of using only the prescribed cultivar 'Agiorgitiko’ , the more productive 'Cabernet Sauvignon’ is sometimes added. DNA samples from fresh and fermented products, containing various mixtures of these two cultivars, were therefore subjected to a CAPS assay . Presence of the adulterant could be detected down to 10% throughout the fermentation process.
Olive oil is also often marketed with PDO labelling. RAPD, ISSR and SSR analysis of Portuguese olive oils allowed the determination of geographic origin of the cultivars on which they had been based . Similarly all 10 olive cultivars involved in samples of Italian oil samples could be identified with only one AFLP primer pair . For rice, the adulteration of the expensive Basmati rice is an important issue, not only for European and US customs but also for consumers. Basmati cultivars have often been mixed with crossbred Basmati varieties and long-grain non-Basmati varieties. Several DNA-based markers have been proposed, and some were commercialized for adulteration tests, such as the multiplexed SSR markers developed by Archak and colleagues . DNA analyses of various plant-based food products have similarly been used for authentication. The presence of the apple 'Annurca’ could thus be verified by SSR analysis in highly processed nectar and purພ products . Using relatively short SSR target sequences (below 160p), it was also possible to amplify genomic DNA from canned pear fruit and fruit juice while markers with longer target sequences failed .
Medicinal drugs constitute another important product area where adulterants cause major problems. Based on nine SNP sites, all populations except two could be distinguished in DNA isolated from the dried stems of the orchid Dendrobium officinale, which is a valuable source of 'Fengdou’ drugs used in traditional Chinese medicine . The latter two populations could instead be distinguished using a more complex procedure known as suppression subtraction hybridization which involves PCR amplification, differential DNA fragment cloning and sequencing. Using these protocols, origination of the plant material could be determined for 50 drug samples obtained at a commercial market. For more information on DNA marker use in medicinal plants, see the reviews by Nybom and Weising  and Sarwat and colleagues .
A variety of DNA marker methods have been used to demonstrate infringement of Plant Breeder’s Rights, either in court or, in our experience much more common, leading to a settlement outside of court . A related field concerns the identification of plants, the possession of which is considered illegal. Thus several studies have been published on the identification of Cannabis sativa specimens as part of drug enforcement . In one approach, 15 SSR loci were combined into a single multiplex to enable fast and user-friendly discrimination between Cannabis genotypes . One of the detected genotypes, however, proved to be very common in police seizure-derived evidence material, suggesting that many illicit growers had access to the same clone. This clonal propagation of course makes it difficult to determine the origination of a particular batch. A related DNA marker application concerns violation of trade restrictions. A special situation is encountered when products from protected trees are involved since woody tissue usually yields heavily degraded DNA. Nevertheless, a set of SNP markers derived from cpDNA intergenic spacers have proven useful for identification of tropical tree species using wood-derived DNA samples .
Genetic diversity, population structure and genetic relatedness
Discrimination among different genotypes is often only a starting point for the subsequent quantification of genetic variability among these genotypes and analysis of patterns of relatedness and gene flow. The extent of genetic variation in a species and its distribution among and within populations is determined by a large number of factors, such as the breeding system, historical events regarding, for example, habitat availability and immigration, population size, migration between populations and many biotic and abiotic ecological factors. Nybom and Bartish  compiled 106 RAPD-based studies and described the effects of several life history characters and sampling strategies on genetic diversity estimates. In another paper , 307 nuclear DNA marker studies (RAPD, AFLP and SSR) were compiled and investigated in a similar manner. One outcome of these surveys was that long-lived, outcrossing and late successional taxa retain most of their variation within populations, whereas annual, selfing and early successional taxa allocate more variation among populations. Within-population diversity is, in general, negatively correlated with the level of population differentiation.
The uniparentally inherited plastid genomes behave as a single, haploid character, and the effective population size for plastid markers is therefore only half of that of nuclear (diploid and biparentally inherited) markers. Consequently, population differentiation due to genetic drift occurs much faster for cpDNA markers than for nuclear markers. Because of their relatively high intra-specific variability, chloroplast and mitochondrial micro- and minisatellites are therefore very useful for studying genetic structure at a species-wide scale.
Population differentiation and gene flow
DNA markers have become a major tool for studying fundamental evolutionary influences of natural selection, mutation, gene flow and genetic drift on wild plant populations. While selection and colonization history is responsible mainly for large-scale structuring of genetic variation, gene flow and genetic drift operate also at a more narrow geographic scale. Among these factors, gene flow especially has received much attention since it is crucial in determining levels of species integrity and subdivision. As already mentioned, breeding system has a profound effect on gene flow and the partitioning of genetic variation between and within populations. The occurrence of IBD between populations has been demonstrated with DNA markers in many different kinds of outcrossing plant species such as, for example, the herb Saxifraga oppositifolia, the Brazilian peppertree Schinus terebinthifolius and the Australian shrub Grevillea mucronulata. IBD has been shown to occur, although much more seldom, also in selfing species such as wild emmer wheat, Triticum dicoccoides. In accordance with these results, a correlation was found between collection distance and RAPD-based among-population diversity estimates for outcrossing taxa . A corresponding association was, however, not found for selfing taxa.
In addition to the inherent dispersal capabilities of a species, gene flow is also affected by natural and anthropogenic habitat heterogeneity. Spatial autocorrelation analysis has thus become a valuable tool for studying spatial scale-dependent changes in DNA marker polymorphism within a population or group of closely occurring populations, and the impact of habitat characteristics on the resulting spatial genetic structure (SGS). Several computational methods have been used to calculate autocorrelation coefficients that measure the genetic similarity between individuals that fall within a defined distance class. A positive autocorrelation is frequently encountered over shorter distances, even if there is no overall linear correlation between geographic and genetic distances when calculated across the whole data set. Using RAPD data, Torres and colleagues  found significant autocorrelation in the first distance class (15 m) in populations of the endangered cliff specialist Antirrhinum microphyllum, suggesting a patchy distribution of genetic diversity. This is consistent with the territorial behavior of the main pollinator Rhodanthidium sticticum, short-distance seed dispersal, and a likewise patchy distribution of suitable habitats.
Many plant species comprise both central, so-called core populations as well as more or less peripheral populations. Such populations may experience considerable differences in the magnitude of operating evolutionary and ecological forces. For example, edge and core populations of the herb Pulmonaria officinalis exhibited strong differences in allelic and genotypic richness, expected heterozygosity and inbreeding coefficent when analyzed with SSR markers . Similarly, an SSR analysis of eastern white cedar, Thuja occidentalis, showed that SGS could be detected over a six times larger distance (90 m) within peripheral populations compared to within core populations (15 m) .
Autocorrelation analysis has demonstrated IBD also in mainly selfing species but then usually at a very narrow scale, as was shown in the wild barley species Hordeum spontaneum.
Highly informative estimations of gene flow can be obtained by genotyping the same plant material using both nuclear and organellar markers. Since the former are biparentally inherited and the latter usually only maternally inherited, the resulting data provide an indication of the relative importance of pollen versus seed migration . This ratio can vary by at least two orders of magnitude, and is typically much lower for insect- as compared to wind-pollinated plants . In dioecious and therefore obligatory outcrossing plants, a mixture of autosomal and sex-linked SSR markers can provide direct evidence of the relative importance of seed versus pollen dispersal. Contrary to previous expectations, similar levels of pollen and seed dispersal were detected in the dioecius perennial plant Silene latifolia. In selfing species, the lower incidence of inter-plant pollen transfer is expected to reduce the pollen to seed migration ratio, as verified by values well below unity at short distances within wild populations of Hordeum spontaneum.
While genes can move between populations by seed and/or pollen, colonization of new habitats is dependent on seed only. In coastal plants, seeds often have the potential to disperse over long distances by hydrochory. In a study of wild sea beet, Beta vulgaris subsp. Maritima, comprising more than a thousand plants from 33 populations along the French coast of the Anglo-Norman gulf, both mitochondrial and nuclear SSRs were applied . Analysis of SGS and determination of zones of sharp genetic change demonstrated narrow IBD indicative of short-range dispersal, as well as genetic barriers fitting the orientation of marine currents and indicative of long-range seed dispersal.
Effects of an increased subdivision or fragmentation of natural plant habitats has received much attention lately dispersal between populations is reduced as well as genetic diversity. Outcrossing species may especially suffer from enforced selfing or biparental inbreeding in fragmented habitats, and lose much of their potential for adaptation to changing environmental conditions. Using SSR markers, White and colleagues  compared fragmented versus continuous populations of the tropical tree Swietenia humilis in Honduras. Genetic variation was still high in all habitat fragments, but low-frequency alleles were more scarce, thus foreboding future genetic erosion. In another early study on tropical trees, Aldrich and Hamrick  reconstructed a population-level pedigree of Symphonia globulifera. Seedlings only occurred in primary and remnant forests, but not in pastures. Surprisingly, however, the majority of seedlings in fragmented forests proved to be derived from a few adult trees located in the open pasture land. Thus the genetic bottleneck experienced by the seedlings in remnant forest patches was caused by the reproductive dominance of a few spatially isolated trees in pasture land, in conjunction with unusually high levels of selfing in these trees.
Overall, tree species have been considered as comparatively resilient to fragmentation due to their often highly effective long-distance dispersal mechanisms. Recently, however, the wind-pollinated and wind-dispersed Andean tree Polylepis multijuga was analyzed with AFLP and shown to contain surprisingly little heterozygosity and to display SGS at short distances, suggesting that most seeds moved only a few meters . This type of information is valuable when developing conservation plans for species protection and perhaps also for a possible reintroduction. Information about, for example, colonization and spreading behavior can be equally helpful when developing measures for stopping further growth of an invasive species. A combination of spatial genetic and geostatistical analyses of data from chloroplast and nuclear SSRs showed how the original two introductions of the invasive Brazilian peppertree Schinus terebinthifolius in western and eastern Florida, respectively, had spread and hybridized in little more than one century . Since both long-distance jumps and short-distance diffusive spread could be demonstrated, highly concerted eradication efforts or the manufacturing of effective biocontrol agents are apparently called for.
DNA fingerprinting data are often used to quantify levels of relatedness among genotypes or groups of genotypes, and numerous relatedness estimators have been described and compared. When wild plants are involved, the purpose is often to compare DNA marker-derived estimations of relatedness with current systematic treatment (see also "Applications of present-day DNA fingerprinting in plants" above). Other applications include parentage analysis, which is the most direct way to estimate gene flow. SSRs are the most commonly used markers for this purpose but data simulations have shown that multi-locus markers such as AFLP can also be used with high confidence, at least when the dominant alleles occur in frequencies of 0.1 to 0.4 . Using SSR, a paternity analysis was conducted in a natural stand with two oak species, Quercus robur and Q. petraea. The spatial distribution of male parents of the offspring from 13 maternal progeny arrays was determined, and the information used for calculation of pollen dispersal curves and analysis of gene flow. Similarly, gene flow was estimated from an SSR-based paternity analysis in the South American palm tree Euterpe edulis. First, an exclusion analysis was performed by comparing adult and juvenile genotypes. After that, a paternity index was calculated among adults that could be the putative parents for a particular juvenile. Gene flow was shown to take place over longer distances than expected (up to 22 km), but it was not possible to distinguish between seed versus pollen transport. Since chloroplasts are paternally inherited in conifers, chloroplast simple sequence repeat (cpSSR) markers can, however, be very useful for direct estimates of paternity, as was demonstrated in white fir, Abies alba.
Access to correctly defined relationships can be very important in plant breeding for the calculation of heritability of specific traits. Various statistical formulae have therefore been developed for determining genetic relationships among individual plants. In a comparison of either purely marker-derived estimations of relationships or combined pedigree and marker-derived estimations, the latter proved to be more informative when analyzing Scots pine, Pinus sylvestris, offspring in a progeny test of open-pollinated genotypes in a seed orchard . Surprisingly incongruent data were obtained when S-SAP markers, SNPs and pedigree data for a set of 35 wheat cultivars were compared . The molecular methods produced similar estimates for the overall partitioning of genetic diversity between and within groups of cultivars, but the genetic similarities between pairs of cultivars were not correlated. SNP-based data were more closely associated with pedigree information than S-SAP-based estimates, probably because polymorphisms are strongly dependent on retrotransposon-related genomic rearrangements.
For cultivated crop plants, estimates of relatedness can provide valuable insights into the domestication process when material originating from different geographic areas is analyzed such as in, for example, Italian olive cultivars . Relatedness among cultivated material on the one hand, and wild populations of the same or closely related species on the other hand, has also been addressed in, for example, apple using SSR, AFLP and cpDNA markers [185,186] (see also "Hybridization and introgression" below).
Estimating the true level of genetic relatedness among cultivars from nuclear DNA marker data is quite difficult, since the obtained information can usually only estimate identity by state (phenetic analysis) instead of the more desirable identity by descent (phylogenetic analysis). One interesting approach towards a true phylogenetic analysis has, however, been achieved within the HiDRAS project . This project involves the analysis of specific chromosomal regions in genetically related apple cultivars using a large set of SSR markers that cover almost the whole genome. Thus, being able to accurately detect levels of genetic relatedness between different cultivars is very helpful for further analyses of, for example, QTL inheritance.
DNA marker-based procedures have frequently been applied to assess diversity and relatedness in collections of cultivated plant material - for example, gene banks. Interestingly, the anticipated loss of overall genetic diversity proved to be negligible when studied in 198 Nordic bread wheat landraces and cultivars that were developed during the last 100 years . DNA markers are also highly useful for the purpose of setting up core collections within gene banks - that is, subsets of the entire plant material, chosen so as to preserve as much as possible of the initial diversity. Two main approaches have been used the first with some kind of stratification using cluster analysis, and the second with methods for determination of genetic uniqueness. Numbers of retained SSR alleles can be maximized using a measure of uniqueness known as maximation strategy [188-190].
Genome constitution: hybridization, introgression and polyploidy
In contrast to animals, many plant groups are characterized by highly variable ploidy levels, often even within the same species. This addition of genomes has certain effects on DNA marker application and data treatment. Moreover, the formation of hybrids by fusion of gametes from two different entities (species, subspecies, and so forth) is also common in plants . While homoploid hybridization takes place at the same ploidy level, most hybridization events instead involve the duplication of genomes, resulting in allopolyploid taxa.
Hybridization and introgression
In a series of classical studies, homoploid hybridization among American Iris species was investigated using a wide variety of DNA-based methods. First, Iris fulva and I. hexagona were shown to each have a species-specific rDNA profile . Subsequently, DNA profiles indicated inter-specific hybridization as well as further introgression in both directions in populations where the two species co-occurred . Diagnostic RAPD and cpDNA-CAPS markers were generated for these two species as well as for I. brevicaulis, and I. nelsonii was shown to have derived from hybridization between all three species .
Another important set of studies on homoploid hybridization has been undertaken in the sunflower genus, Helianthus. RAPD linkage maps were developed for the sympatric and hybridizing species H. petiolaris and H. annuus and subsequently used to analyze the genome of a recently formed hybrid species, H. anomalus, as well as of an artificially generated hybrid [191,195]. Later on, divergence between the two parental species was analyzed using 108 mapped SSR markers , and below average introgression was noted for SSR markers located close to QTLs for species differences when two parapatric species, H. annuus and H. debilis, were investigated [197,198]. Interestingly, gene flow was mainly in the direction from the hybrid back into these two parental species .
cpDNA-derived information has played a major role in elucidating many cases of homoploid hybridization and subsequent introgression. Studies of multiple taxa in several tree genera have thus shown that chloroplast haplotypes often are closer associated with geographic origin than with species affiliation - for example, in oak trees, Quercus, in Eucalyptus from Tasmania  and in the South East Asian pioneer tree genus Macaranga. This introgression phenomenon has been coined 𠇌hloroplast capture” .
DNA markers are also commonly used for detecting both ancient and ongoing hybridization between crops and their wild relatives. Malus sieversii grows in Kazakhstan and has been suggested as progenitor of cultivated apple, M. domestica, based on morphological, historical and molecular evidence . Nuclear SSR-based analyses have later been undertaken to investigate the genetic diversity and population structure in M. sieversii. The origination of cultivated apple may, however, be more complicated. In another SSR-based study, three separate although partly overlapping gene pools were formed by (1) M. sieversii, (2) the European wild apple species M. silvestris, and (3) old and modern apple cultivars . In the same plant material, analyses of chloroplast haplotypes produced rather unexpected results. Thus, M. sylvestris not only had the same common haplotypes as M. domestica, but there was also local sharing of uncommon haplotypes between the two species, suggesting recent inter-specific gene flow. A strong affinity between M. sylvestris and modern apple cultivars was likewise suggested in an SSR analysis of 839 genotypes collected from China to Spain, and representing four wild species as well as cultivated apple . In this study, data were analyzed both with the computer program STRUCTURE, and with approximate Bayesian computation which offers a more historical perspective on gene flow.
Two variants of STRUCTURE, InStruct and NewHybrids, were used by Muranishi and colleagues  in a recent SSR-based study of Magnolia stellata and M. salicifolia together with putative F1 and F2 hybrids and backcrosses. The resulting clusters could be verified also with morphological trait analysis. Simultaneous application of cpDNA SSR markers showed that introgression was heavily asymmetric, with M. salicifolia being the seed parent of almost all hybrids and backcrosses.
In plant breeding, there can be good reasons for analyzing parental contributions in recently developed, experimental hybrids, especially if the breeding process has involved one or several generations of backcrossing. The amount of parental influence could thus be quantified using a microarray analysis with 7,680 probes simultaneously detecting SNPs, indels and methylation differences, in a set of intergeneric hybrids between the commercially important grass genera Festuca and Lolium. The extent of similarity between the derived Festulolium cultivars and the parental genomes was clearly associated with the type of crossings performed - that is, F1, F2 or backcrosses.
Polyploidy is very common in the plant kingdom. Although the same marker technologies can be used for genotyping diploid as well as polyploid samples, statistical analyses and interpretations are usually less straightforward when polyploid samples are involved. Many species are allopolyploids and have been derived from their diploid ancestors by hybridization. Moreover, molecular studies of allopolyploid taxa and their putative progenitor taxa have shown that multiple origination is the rule rather than the exception. While frequent gene flow between polyploid lineages and back-crossing to parental taxa can further confound this process, a more easily studied case is offered in apomictic species where the speciation event is more or less frozen in time. One such example is the North American allopolyploid cloak fern, Astrolepis integerrima, which was recently studied by cpDNA sequencing and AFLP analysis . Six relatively localized cpDNA haplotypes were detected, some of which were further divided by AFLP. All in all, the results suggested that a total of 10 A. integerrima lineages have been formed through multiple independent hybridizations between A. cochisensis and A. obscura.
Identification of the putative progenitor species of polyploids can be attempted with various types of markers, including the internal transcribed spacer (ITS) region in the nuclear ribosomal RNA gene clusters which was sequenced and analyzed in, for example, polyploid rose cultivars and species . Nair and colleagues  used IRAP primers to determine the genomic constitution in a set of mostly triploid banana cultivars. Primer sequences were derived from two different retrotransposons, one occurring in the A genome (Musa acuminata) and the other in the B genome (M. balbisiana). A more easily applied CAPS marker, obtained from PCR amplification of the ITS region followed by restriction with RsaI, has also been applied for this task . More recently, application of 653 DArT markers similarly allowed the discrimination between A and B genomes, and the identification of these genomes within a set of banana cultivars .
Multi-locus based methods such as RAPD and AFLP are sometimes used for studying population genetics in species with different ploidy levels, but problems can arise due to a positive correlation between ploidy level and number of scored bands . In addition, banding patterns may differ qualitatively between samples at different ploidy levels, and thus give rise to scoring errors. Single-locus markers such as SSRs and SNPs are also problematic due to the occurrence of multiple alleles and complex segregation ratios. SSR markers may be more or less genome (and species)-specific and therefore fit only one of the two homologous genomes of an allopolyploid hybrid, producing no amplification in the other (null alleles, or allele drop-out). Truly genome-specific SSR loci that consistently produce a maximum of only two alleles in each sample are rare, but can be quite useful as demonstrated in the hexaploid Mercurialis annua. With these markers, population genetics parameters could be calculated as if the species instead were diploid.
For allopolyploid crops with intermediate levels of similarity among homologous genomes, such as tetraploid potato, SSR primers in general produce a variable number of bands per locus. For example, Fu and colleagues  found a total of 64 alleles when investigating 169 potato accessions with 36 SSR primer pairs. Even apparently diploid species such as apple may be 𠇊ncient polyploids” in which some primer pairs can produce a second set of alleles derived from an unrecognized duplicated genomic area . Amplification of these supernumerary loci (isoloci) frequently varies with the experimental conditions, and can cause problems when data are being combined from several laboratories.
In autopolyploids and in allopolyploids with low genomic differentiation, SSR analyses usually produce multiple alleles of a single locus in each genotype, as demonstrated in the autopolyploid and apomictic Ranunculus kuepferi and in allopolyploid species and cultivars in the genus Rosa[217-219]. To fully utilize the information content of the obtained DNA profiles, segregation patterns must be determined. This, however, requires the ability to score allele dosage, in contrast to just the presence or absence of an allele. The MAC-PR approach (microsatellite DNA allele counting - peak ratios) determines allele copy number based on quantitative differences between microsatellite allele peak ratios and therefore allows the precise determination of allelic configuration in each studied sample, as was shown in tetraploid roses [217,220]. Using this approach, inheritance patterns have been studied even in the absence of experimental crosses . High-quality banding patterns are, however, needed for successful application of the MAC-PR, as well as repeatability of relative allelic amplification intensities among individuals and, thus, homology of microsatellite marker alleles within a species.
In crop plants with detailed pedigree information, the so-called microsatellite allele dose and configuration establishment (MADCE) procedure can be used to trace the transmittal of SSR alleles through documented generations of the investigated plant material, and determine the exact allele copy number in the target cultivars . Originally, the MADCE procedure was applied in apple  but informative results have recently been obtained also for the Strawberry Crop Reference Set within the RosBreed research project (Bassil N, personal communication).
In many situations, allele dosage can, however, not be accurately scored with the methods chosen, and much plant material lacks or has only unsatisfactory pedigree information. Specialized programs have therefore been developed for analyzing polyploids with SSR markers, such as, for example, TETRA  and POLYSAT . The fitTetra R package has been developed for enabling genotype calling in tetraploid species from biallelic marker data, and is especially useful for large-scale SNP analyses in material with high levels of polysomic allele segregation such as potato . By contrast, the bead array MSV package  appears to be more useful for material with mainly disomic segregation.
A major drawback with any multi-locus approach is the loss of information about exact levels of heterozygosity and about genome inheritance. In addition, genetic distances between cultivars are exaggerated as compared to distances calculated on the basis of co-dominant data . A method for calculating genetic distances that permits unbiased comparisons between different ploidy levels has, however, been described  and is available in the computer program package GENOTYPE/GENODIVE. Another approach is based on the formation of multi-locus allele phenotypes of each investigated individual, and calculation of phenotype-based estimates of genetic diversity and differentiation . For more information on methods to describe the population genetics of polyploids, see Assoumane and colleagues .
Plant speciation, phylogeny and systematics
As the availability of DNA-based information increases, more attention is being paid to the genomic patterns of differentiation among plant species. According to the genic view of plant speciation, small “genomic islands” may be responsible for much of the differentiation between taxa through divergent selection or reproductive isolation barriers, while the remainder so-called “porous genome” is more permeable to gene flow [230,231]. In this context, the choice of molecular method becomes crucial for the ability to reflect genomic differentiation in a phylogenetically relevant perspective. To determine relative marker sensitivity in monitoring inter-specific differentiation, Scotti-Saintagne and colleagues  conducted a genome scanning experiment with 389 markers (allozymes, AFLPs, SCARs, SSRs and SNPs) on samples from pairs of populations of the sympatric oak species Quercus robur and Q. petraea. Distribution of markers according to their ability to detect inter-species diversity was clearly L-shaped apparently only a few markers were located in genomic regions responsible for species differentiation. As expected, these markers were more likely to reside in coding regions than in non-coding regions. In another genome scan based on 88 mapped SSR loci, most loci again showed considerable migration between the analyzed taxa: sunflower species Helianthus annuus, H. debilis and their inter-specific hybrid . The genomic regions that are responsible for genetic differentiation therefore appear to be small in these taxa, whether estimated as level of species differentiation or as migration rates.
When targeting differentiation at a larger taxonomic scale, DNA sequence information of the chloroplast genome and/or nuclear genic regions, is usually preferred over DNA fingerprinting. Proper elucidation of the complex puzzle of plant systematics is, however, often best achieved with a combination of different types of molecular information. In many plant groups, various types of multi-locus or single-locus markers have thus provided important pieces to the puzzle. So far, these marker loci have seldom been placed on a genomic map (but see, for example, [199,232]), and whether they reside in a conserved versus a “porous” part of the genome is usually unknown. Instead, choice of markers has mainly been based on the feasibility of obtaining a sufficiently large number of polymorphic bands.
Multi-locus DNA profiling methods such as AFLPs have become the most commonly used DNA fingerprinting tool in plant systematics, mainly in situations where DNA sequencing produces insufficient phylogenetic resolution . In an early plant systematic study using AFLPs, 551 polymorphic bands were obtained with three primer combinations for 30 accessions from 19 taxa of Solanum section Petota and three taxa of Solanum section Lycopersicum. Ploidy level was reflected in the profiles, with hexaploids exhibiting more bands than tetraploids and diploids. Mating system had, as expected, a large impact, with 40 to 60% intra-specific polymorphism detected in outcrossing taxa as compared to only 0 to 2% in selfing taxa. AFLP methodology was also employed to investigate phylogenetic relationships among 43 species of the paleotropic pioneer tree genus Macaranga. About 30 of these species have a symbiotic relationship with specific ant partners. The resulting phenograms supported the monophyly of several sections and subsectional groups within the genus, and provided evidence for a polyphyletic origin of the ant-plant mutualism.
Besides the two species of cultivated rice, the genus Oryza also comprises of around 22 wild species that have received considerable attention due to their potential importance for rice breeding. Six diploid genomes (A, B, C, E, F and G) and four allotetraploids (BC, CD, HJ and HK) have been identified using, among other methods, total genomic DNA hybridization . In another early study on rice, 77 samples representing 23 Oryza species were analyzed with AFLP . Pairwise genetic distances showed a linear increase depending on the taxonomic level, with 0.02 to 0.21 within species, 0.2 to 0.35 between species sharing the same genome type, and Ϡ.7 between species carrying different genomes. For the subsequent analysis of phylogenetic relationships among these genomes, more conserved markers were developed through the identification and sequencing of numerous rice genes . Comparison of sequences for 142 such genes in six species, representing the six different diploid genomes, allowed the reconstruction of the rapid diversification in Oryza. In a follow-up study based on the sequences of 106 nuclear genes, divergence times and ancestral effective population sizes were also determined .
In the large and complex genus Rosa, several different DNA-based methods have been applied for phenetic and phylogenetic analyses, with mostly consistent results - for example SSR  and AFLP . Two major clades were identified, with sections Carolinae, Cinnamomeae and parts of Pimpinellifoliae forming one clade and most of the other seven commonly recognized sections forming the other clade. As for the division into sections, Synstylae appears to be mainly monophyletic and rather closely affiliated with sections Indicae and Rosa. Furthermore, section Pimpinellifoliae is apparently polyphyletic, and R. spinosissima should be separated from the other species in this section. In spite of its size (currently, about 50 species are acknowledged) and hybridogenous origination, the mainly European section Caninae (also known as dog roses) apparently constitutes a well-circumscribed monophyletic group. Another, very large AFLP study was recently conducted on dog rose specimens sampled in a transect across Europe, with more than 200 non-dog rose samples analyzed for comparison . Two lines of statistical analyses were applied: (1) an unstructured model with principal coordinate analysis and hierarchical clustering, and (2) a model with a superimposed taxonomic structure based on analysis of genetic diversity using a novel approach that combines assignment tests with canonical discriminant analysis. Support was found for five of the seven subsections, including the three major ones: Caninae, Rubigineae and Vestitae. Within the subsections, many species overlapped considerably, and geographic distances often appeared to be at least as important as the conventional taxonomy in explaining similarities between analyzed specimens. Complementary information on phylogeny in Rosa has also been obtained with DNA sequencing. Although sharing some ITS sequence types with species in other sections thereby confirming their hybridogenous origin, the Caninae species also have one unique ITS sequence type which is further evidence of their monophyly [209,242].
Plant systematic studies have occasionally been conducted also using SSR markers, especially when the focus has been on genetic differentiation among closely related taxa. For example, ten Puerto Rican populations of the cycad genus Zamia were analyzed with 31 SSR primer pairs . These populations could be treated either as belonging to a single polymorphic species, Z. pumila, or as representing three more narrowly circumscribed taxa: Z. erosa, Z. portoricensis and Z. pumila sensu stricto. The SSR analysis showed that Z. erosa is strongly differentiated from the other two species, and thus may represent an independent introduction into Puerto Rico. The data are consistent with an allopatric speciation scenario with Z. portoricensis being the youngest taxon according to Bayesian coalescent analysis and effective population size, and still showing considerable admixture with Z. pumila.
Genetic relationships among 35 Arachis species from seven sections, including 11 accessions of cultivated peanut, A. hypogaea, were analyzed on the basis of allelic variation at 32 SSR loci . A neighbor joining tree was generated on the basis of pairwise Dice distances between individual accessions, calculated from a binary presence/absence matrix of SSR alleles. Most con-specific accessions grouped together on the tree, as did species from the same section, with several exceptions that were attributed by the authors to either homoplasy in the dataset or extensive within-species variation.
From the relatively few studies available, it appears that SSR markers do have some potential not only for species delimitation, but also for the reconstruction of genetic relationships among closely related species groups that are only a few million years old. However, SSR markers are usually highly polymorphic and therefore multiallelic within a species. Accordingly, the within-population component of SSR variation is often much higher than the between-population or between-species component . It is therefore a “must” that several accessions per species are included in any phylogenetic study that is based on SSRs, the more the better. Optimally, genetic distances between populations (or species) rather than genetic distances between individuals should be used to generate phenetic trees.
Phylogeography aims to study the spatio-temporal history of a species on the basis of its intra-specific genetic variation . In principle, phylogeographic studies can be based on information from either nuclear, mitochondrial or chloroplast DNA. In practice, organellar DNA is usually preferred since organelle-derived markers are more likely to retain information about biogeographical history than nuclear markers . There are several reasons for this. First, the haploid genomes of plastids and mitochondria exhibit a smaller effective population size as compared with the diploid nuclear genome, resulting in stronger substructuring of fragmented populations under genetic drift. Second, organellar genomes are usually inherited uniparentally. In angiosperms, the plastid DNA is generally transmitted by seeds - that is, maternally. Given that plants can colonize a new habitat only by seeds, plastid-derived markers have the potential to provide information about past changes in species distribution that is unaffected by pollen flow. Third, intermolecular recombination is usually absent in plastid DNA, so that individual sequence polymorphisms can be combined into haplotypes that remain mostly unchanged when passed to the next generation.
Evolutionary relationships between cpDNA haplotypes are often depicted as networks , which can be superimposed on the geographic distribution of the sampled plants. One has to keep in mind, however, that a non-recombining DNA molecule behaves like a single gene. The phylogeographic pattern retrieved from a plastid haplotype network therefore only represents one out of several possible outcomes of the genealogical process . This is why phylogeographic analyses based on other genes and genomes are becoming increasingly popular. In conifers, where plastid DNA (paternal) and mitochondrial DNA (maternal) show contrasting modes of transmission from parents to offspring, both genomes have often been analyzed side-by-side [250,251]. In addition, phylogeographic studies often employ nuclear ribosomal ITS sequences. In most plant species, the ribosomal genes are rapidly homogenized by concerted evolution and then behave like uniparentally inherited organellar DNA.
There is no clear division between phylogeography on the one hand and traditional population genetics on the other. Accordingly, the use of nuclear SSR markers to study genetic diversity, genetic subdivision and gene flow within and among extant species is sometimes also called “phylogeography” , and there have been numerous successful attempts to elucidate intra-specific phylogeographic patterns by multi-locus DNA profiling methods such as RAPD, ISSR and AFLP [253,254]. Multi-locus banding patterns are typically analyzed phenetically - that is, phenograms or networks are reconstructed on the basis of a pair-wise similarity matrix that is generated from a binary presence/absence matrix of band positions. The (groups of) genotypes depicted in the resulting phenogram or network are then compared with their geographic distribution .
The majority of plant phylogeographic studies still rely on plastid DNA polymorphisms that can be searched for by either PCR-RFLP, screening of length-variable plastid microsatellites (cpSSRs), or by comparative sequencing of PCR-amplified non-coding DNA . Unique polymorphisms are then combined into distinct haplotypes, followed by the analysis of haplotype distribution and frequencies in different geographical regions, quantification of the genetic divergence between haplotypes, and the evaluation of genetic relationships between haplotypes - for example, in the form of a statistical parsimony network such as TCS . The use of cpSSRs is, however, controversial, since their often high mutation rates can cause homoplasy [257,258].
Application areas of marker-based phylogeographic studies are diverse, and include, for example, the analysis of postglacial re-colonization patterns of the Central European landscape by trees and shrubs in the Quaternary [252,259], the identification of glacial refugia , the reconstruction of migration routes of halophytes along coastal and inland salty habitats , the investigation of the evolutionary history of tropical trees  and the historical biogeography of threatened species . Increasingly important are comparative phylogeographies that involve numerous animals and plants from the same geographical region [261,262]. Such meta-analyses yield invaluable data on common evolutionary patterns across many biota from large geographical areas.
Linkage mapping and genetic maps
One prominent application of molecular markers is the generation of genetic maps which have been established for all major and many minor crops and other plants (for example, rice , barley , and maize , to name just a few). A genetic map is a graphic representation of a chromosome (or linkage group) onto which genetic elements (= loci, for example markers or genes) are aligned. The loci are arranged based on their co-segregation during meiosis, which depends on the frequency of recombination events. Genetic distances between loci are measured in centiMorgan (cM). One cM is defined as the distance that two loci have to each other, if in 100 meiotic events the loci are segregating only once (= 99% chance of co-segregation). As the extent of recombination varies in different genomes, this translates into varying physical distances. The recombination frequency also varies among different genomic regions - for example, recombination is suppressed near centromeres.
To estimate genetic distances among loci, the co-segregation of genetic elements is monitored in mapping populations or in association mapping approaches (see below). Mapping populations usually originate from a cross between two parental lines, which ideally can be distinguished by a large number of polymorphisms that are monitored in the progeny. Particularly convenient mapping populations consist of so-called Recombinant Inbred Lines (RILs), which are generated by selfing single-seed descent from different sibling F2 plants through six or more generations. The continuous selfing causes very high levels of homozygosity, and each RIL from a population of RILs hence conserves one particular recombinational event from the F1 cross. Design and construction of RILs have been reviewed by Pollard .
To identify loci that are very tightly linked with a specific trait, 𠇏ine-mapping” is performed by enriching the density of markers in proximity of the responsible genes or, in the best case, markers for the responsible genetic elements themselves. The most commonly applied technique for fine-mapping is Bulked Segregant Analysis (BSA), originally developed by Michelmore and colleagues . In a BSA, all genotypes that show a specific phenotype (that is, a specific trait) are pooled and screened for polymorphisms that distinguish them from the remaining plants. All genetic elements that do not influence the bulk-trait are randomly distributed among all plants, whereas all genetic elements responsible for the trait are to be found preferentially if not only in the respective bulk. In consequence, any difference between the bulk and the remaining plants is likely to be linked with the trait of interest. The source of polymorphisms can be, for example, the metabolome, the proteome, the transcriptome or the genome. The latter two have profited enormously from the advent of high-throughput-sequencing technologies and are now the most widely used sources for genetic polymorphisms.
Different types of markers can be combined into integrated maps, which become more highly resolved (that is, saturated) with each newly added marker. Furthermore, data from different crosses can be integrated in the same map. For example, Wenzl and colleagues  published an integrated map for barley using DArT, SSR, RFLP and STS markers, altogether comprising 2,935 different loci. In the current era of genome sequencing, genetic maps are also a versatile tool for defining the order of assembled contigs from shotgun sequencing approaches, as has been done, for example, during the assembly of the recently published chickpea genome .
Association mapping (AM) aims at linking phenotypes to genotypes, independent of the kinship of the genotypes. The concept of AM has been implemented in humans and model organisms for many years (for example within the human HapMap project that started in 2002), and is now increasingly applied for plant genomes (see the reviews by Abdurakhmonov and Abdukarimov  and Soto-Cerda and Cloutier ). The major advantage of AM over linkage mapping (LM) or QTL mapping is that no mapping population is required. The establishment of good mapping populations is a time-consuming and costly task, especially for plants with long generation times and hence a limited number of meiotic recombinations. Furthermore, LM is usually restricted to a small subset of genotypes and to those loci that are polymorphic among these genotypes (= low allelic richness). In contrast, AM examines genotype-phenotype correlations in a large germplasm and hence monitors the historical meiotic recombination events that accumulated in natural populations and collections of landraces, breeding materials and varieties .
Association mapping is based on the occurrence of Linkage Disequilibrium (LD) between a particular trait and one or more alleles of a marker locus in a population. In contrast to LM, which refers to the combined inheritance of loci due to their close physical proximity on the same chromosome, LD refers to the non-random occurrence of allele combinations of loci in a population. Thus, the reason for LD can be linkage (and in most cases it is), but also other factors influence LD, such as selection, mutation, mating system, population structure, and so forth, which can result in significant LD even of alleles that are located on different chromosomes . Because of this, AM is more complex than linkage mapping and might be biased by various factors.
More recently, AM-based analyses have been successfully carried out in many crops. In rice, for example, Zhao and colleagues  genotyped more than 44,000 SNPs across 413 accessions from 82 countries. Dozens of variants could be identified that influence numerous complex traits. In maize, a high-density analysis based on 56,110 SNPs was performed to analyze chilling tolerance in 375 inbred lines . Nineteen highly significant association signals that explained between 5.7 and 52.5% of the phenotypic variance observed for early growth and chlorophyll fluorescence parameters were identified. An AM-based approach that was termed “landscape genomics” aims at simultaneously examining the effects of demographic history, migration and selection in a defined geographical site (see, for example, Sork and colleagues and references cited therein ).
Thanks to new, high-density genotyping methods such as the INfinumHD assay that assesses thousands of markers simultaneously, almost all major crops can now be subjected to AM . High-density SNP arrays that comprise the information of several thousand loci were recently also developed for forest trees and horticultural plants, including white spruce, Picea glauca, peach , apple  and sweet and sour cherry , and are expected to greatly facilitate AM also in these plant species. A combination of LM and AM represents a particularly powerful tool for selection. Thus, Yu and colleagues  presented a so-called “nested association mapping” (NAS) approach in maize that involved crossing of 25 different variants and 5,000 offspring, whereas Kover and colleagues  performed a “Multiparent Advanced Generation Inter-Cross” in Arabidopsis. Nineteen variants were crossed in a random mating scheme, resulting in 527ꃴ plants. The RILs originating from the maize NAS recently helped to identify important genes involved in maize kernel composition , resistance to northern leaf blight  and stalk strength . In apple, Khan and colleagues  used an AM and LM combined approach to identify three important QTLs for fire blight resistance.
Marker-assisted breeding and genome-wide selection
One major aim of genetic linkage analysis in crop plants is marker-assisted breeding (see the review by Jiang ). A particularly promising current concept of marker-assisted breeding has been termed “genomic selection” (GS) or “genome-wide selection” (GWS) . In contrast to the traditional marker assisted selection (MAS) concept, where only a subset of markers is considered, in GS all available markers are evaluated simultaneously for the calculation of a so-called breeding value. This is done by combining major and minor QTLs according to Meuwissen and colleagues . In this way, QTLs with only minor positive and negative effects that are missed in traditional MAS are also taken into consideration for selection. The concept of GWS is widely used in livestock breeding and has been discussed as a future selection approach also for plants . In their analysis that was based on a large data set of 25 nested association mapping populations, Guo and colleagues  found better predictions using the GWS approach as compared with MAS for flowering traits in maize (days to silking, days to anthesis and anthesis-silking interval). MAS was performed by composite interval mapping (see the review by Zou and Zeng ), and GS using a “ridge regression-best linear unbiased prediction” to calculate breeding values.
For genotyping moderate numbers of SNP loci in hundreds to thousands of samples, PCR-based approaches provide a more flexible alternative to microarray-based methods. Besides direct sequencing, three currently popular methods for SNP genotyping of PCR products are high-resolution melting (HRM) analysis, allele-specific PCR (ASP), and the TaqMan assay. In HRM analysis, the PCR product is continuously heated, and the separation of the two DNA strands is monitored in real time . Polymorphic PCR products that differ slightly in length or sequence will have different melting temperatures. These differences can be measured with sensitive optics, which monitor the fluorescence-to-signal intensity of an intercalating dye. HRM analysis has, for example, been successfully applied to gene mapping in rice , to cultivar identification in sweet cherry , and to discriminate between closely related chloroplast DNA haplotypes in the wild species Arenaria ciliata and A. norvegica.
In ASP, the alternative alleles at a particular polymorphic site are amplified with allele-specific primers that are each labelled with a different fluorochrome. The presence of a particular allele is hence indicated by a diagnostic fluorochrome signal. If primers of different length are used, the ASP products can also be assessed by gel electrophoresis. The TaqMan™ assay dates back to the early 1990s . It involves the fluorescence-based detection and quantification of a specific probe that is hybridized to the SNP site of interest. A light signal is only emitted when the probe is degraded by the exonuclease activity of the Taq DNA polymerase, which occurs only when the probe has specifically bound to its target site. The ASP and TaqMan™ assays can be assessed in regular quantitative PCR machines while HRM requires specific optics.
The status of traditional DNA fingerprinting: concluding remarks
Taken together, the last two decades have witnessed a prominent increase in the application of various DNA markers for plant DNA fingerprinting. In the beginning, multi-locus dominant markers, especially RFLP, RAPD, AFLP and ISSR, were most popular, but single-locus SSRs and eventually SNP markers rapidly caught up. The chip-based DArT technology is also still used. We believe that traditional multi-locus methods and their various spin-offs will still be employed a decade from now, but mostly for exploratory research that does not necessarily result in published papers. Locus-specific SSR markers will probably remain more popular, due to their co-dominant inheritance, ease of analysis and the fact that new data can easily be added to already existing files. Novel input to DNA fingerprinting was, however, provided by an unforeseen major breakthrough in DNA sequencing technology, which can be envisaged as the starting point for the future of DNA fingerprinting, and which is discussed in the following section.
Many aspects of differentiation are controlled by hormones . The hormone auxin, for example, plays an important role in the differentiation of vessel elements, both in intact and wounded plants. This role was first demonstrated in experiments where small incisions were made in stem internodes that cut though the phloem and xylem of a single vascular bundle. Auxin produced by the apical meristem and young leaves above the wound induces parenchyma cells to regenerate the damaged vascular tissue. Parenchyma cells undergo transdifferentiation.
Although they already had differentiated as parenchyma cells from ground meristem precursors, they now repeat the steps that procambial cells take when they differentiate as vessel elements. Cells are induced to do this in a chainlike pattern, so that a new continuous strand of vascular tissue is formed as a detour around the original incision. Scientists know that auxin is involved, since transdifferentiation is blocked when the sources of natural auxin (young leaves and buds) are removed or when auxin transport inhibitors are applied. If natural sources of auxin are removed, and artificial sources added, transdifferentiation of parenchyma cells will occur, regenerating the vascular bundle.
Agosta, William. Bombardier Beetles and Fever Trees: A Close-up Look at Chemical Warfare and Signals in Animals and Plants. Reading, MA: Addison-Wesley, 1996.
Bidlack, Wayne R. Phytochemicals as Bioactive Agents. Lancaster, PA: Technomic Publishers, 2000.
Karban, Richard, and Ian T. Baldwin. Induced Responses to Herbivory. Chicago: University of Chicago Press, 1997.
Rosenthal, Gerald A., and May R. Berenbaum. Herbivores, Their Interactions with Secondary Plant Metabolites. San Diego, CA: Academic Press, 1991.
The use of PEG and Mg 2+ instead of Spd and Ca 2+ in the conventional biolistic coating procedure was developed to transform wheat and other plant species in both transient and stable transformation. In bombarding a low quantity of minimal DNA cassette was reported to efficiently regenerate low copy transgenic plants. Nanogram amounts of the minimal expression cassettes of the GOI and the hpt gene were routinely used in high-throughput experiments to generate single copy transgenic plants of commercial wheat at a high frequency.