Author version

AUTHOR VERSION

6

An Overview of Peanut Genome Structure

Bertioli DJ¹, Araujo ACG²*, Nielen S³, Heslop-Harrison P⁴, Guimarães PM⁵, Schwarzacher T⁶, Isobe S⁷, Shirasawa K⁸, Leal-Bertioli SCM⁹

¹Institute of Biological Sciences, Genetics Department, University of Brasilia, Brazil, 70910-900 Brasília-DF, Brazil.

[e-mail]; Tel: [phone]; Fax: [phone]

²Embrapa Genetic Resources and Biotechnology. PqEB W5 Norte Final, 70770-917 Brasília-DF, Brazil. CEP 71.770-900.

[e-mail]; Tel: [phone]; Fax: [phone]

³Plant Breeding and Genetics Section, International Atomic Energy Agency, Vienna, Austria. [e-mail]

⁴Department of Biology, University of Leicester, Leicester, LE1 7RH, UK.

Email: [e-mail]; Tel: [phone] / 3381; Fax: [phone]

⁵Embrapa Genetic Resources and Biotechnology. PqEB W5 Norte Final, 70770-917 Brasília-DF, Brazil. CEP 71.770-900.

[e-mail]; Tel: [phone]; Fax: [phone]

⁶Department of Biology, University of Leicester, Leicester, LE1 7RH, UK.

Email: [e-mail]; Tel: [phone] / 3381; Fax: [phone]

⁷Department of Plant Genome Research, Kazusa DNA Research Institute, Chiba, Japan.

[e-mail]; Tel: [phone]; Fax: [phone]

⁸Department of Plant Genome Research, Kazusa DNA Research Institute, Chiba, Japan.

[e-mail]; Tel: [phone]; Fax: [phone]

⁹Embrapa Genetic Resources and Biotechnology. PqEB W5 Norte Final, 70770-917 Brasília-DF, Brazil. CEP 71.770-900.

[e-mail]; Tel: [phone]; Fax: [phone]

*Corresponding author: [e-mail]

Abstract

Cultivated peanut (Arachis hypogaea L.) is a Papilionoid grain legume crop, important throughout the tropics. It is an allotetraploid of recent origin with an AB type genome (2n = 4x = 40) and has very low DNA polymorphism, a characteristic that has hampered genetic studies. The A and B genomes are of similar size and are composed mostly of metacentric chromosomes. The A genome is characterized by a pair of small chromosomes and the presence of strong centromeric heterochromatic bands, in contrast, B chromosomes are all of similar size and have much weaker centromeric bands. The genome of peanut is estimated at about 2.8 Gb and with a high repetitive DNA content. Its most probable diploid ancestors are A. duranensis and A. ipaënsis, donors of the A and B genomes, respectively. These two subgenomes diverged from a common ancestor about three and a half million years ago, more recently than the subgenomes of cotton or soybean. Consequently, homeologous A and B genic sequences have very high sequence identity. Genetically, cultivated peanut behaves as a diploid. The two subgenomes have very high genetic synteny, and do not appear to have undergone major structural rearrangements after polyploidization. Indeed, the peanut subgenomes even have detectable genetic synteny with other legumes that diverged during evolution about 55 million years ago. The patterns of synteny indicate that the A and B genomes are diploidized, and that gene-space is likely to be ordered into about ten conserved blocks. In contrast to their conserved genetic synteny, the repetitive DNA components of the subgenomes are very significantly diverged. This may be substantially explained by the activity of a few retrotransposons since the time of genome divergence. In addition, the peanut genome harbors many miniature inverted-repeat transposable elements that have been active since polyploidization. This activity has probably contributed to the phenotypic variability of peanut.

1. 
   Introduction
   

It may seem a difficult task to write about peanut genome structure when the sequencing of the entire genome is imminent. At present, we know only a very small amount of DNA sequence and what is known is largely unordered and unanchored. What, the reader may ask, could there be any significance to write that will not be hopelessly outdated in a couple of years? Fortunately however, many broad features of the peanut genome are already apparent, and the knowledge of these features will be of help to put into context and even to generate, assemble and order an entire genome sequence.

Many characteristics of a genome can be observed at higher, or more macro levels than DNA sequencing. Cytogenetics, genetics and linkage mapping are prime examples that have given useful information on the large-scale organization of the Arachis genome. In addition, a relatively small amount of DNA information is very informative of genome structure because, in eukaryotes, typically a few very abundant transposons make up a substantial proportion of the genome. Already, some of the most important properties of the transposon components of the peanut genome are apparent, and some of the most abundant transposons are now well defined.

1. 
   The Position of the Genus Arachis within the Legumes and Base Chromosome Number
   

Peanut, like most other economically important legumes is within the subfamily Papilionoid. Most of these important legumes fall within two subclades of this subfamily that diverged from each other about 50 Million years ago (Mya), the Phaseoloids and Galegoids (Lewis et al. 1995; Lavin et al. 2005). The Phaseoloids, which include soya and common bean, have a base chromosome number of 2x = 20 or 22. The Galegoids, which include pea, lentil, Lotus and clover have a base chromosome number of 2x = 10-16. However, Arachis falls in a different clade, the Dalbergioids. This clade diverged from the Galegoids and Phaseoloids about 55 Mya (Cronk et al. 2006). Dalbergioids are predominantly New World and tropical and have an ancestral chromosome number of 2x = 20. Accordingly, most species of Arachis have 20 chromosomes and those that have different chromosome numbers can be assumed to be derived states. Of particular interest here is cultivated peanut that differs from most of its wild relatives in having an allotetraploid genome (2n = 4x = 40).

2. 
   The Chromosomes of Wild Arachis and Cultivated Peanut
   

Peanut has mostly metacentric chromosomes of similar size. The diploid wild Arachis species that are most closely related to it, and that have symmetrical karyotypes, have been assigned to two genome types, A and B, and belong to the botanical section Arachis. Species with the A genome have a small pair of chromosomes, “the A chromosomes” (Husted 1936; Smartt et al. 1978). Those species with symmetric karyotypes but without A chromosomes were traditionally considered B genome species more properly called non-A species. Recently, based on heterochromatin distribution and rDNA loci, these non-A species have been divided into three groups: B sensu stricto, F and K (Seijo et al. 2004; Robledo and Seijo 2010). The B genome sensu stricto has undetectable or much weaker centromeric heterochromatin, whilst F and K genomes have heterochromatic bands on most chromosomes, but differ in the amount and distribution. Phylogenies based on DNA sequence data strongly support the validity of this division (Moretzsohn et al. 2004; Milla et al. 2005; Tallury et al. 2005; Bravo et al. 2006; Bechara et al. 2010).

Cytological analysis showing heterochromatic distribution, rDNA loci and genome in situ hybridization (GISH) indicated that the karyotype of cultivated peanut is equivalent to the sum of the karyotypes of A. duranensis (A genome) and A. ipaënsis (B genome), or very closely related species (Seijo et al. 2004, 2007). This origin is supported by DNA sequence data, the analysis of which has also revealed a very limited genetic variability in peanut and the wild tetraploid A. monticola (Halward et al. 1991; Kochert et al. 1996; Raina et al. 2001; Milla et al. 2005). Indeed, these two species are most probably the same species. This very limited DNA variability also indicates that A. hypogaea and A. monticola most probably had their origin in a single or very few hybridizations followed by chromosome duplication. It is not known if this origin occurred in the wild, or spontaneously when the two diploids were cultivated in close proximity by ancient inhabitants of South America. In either case, archaeological studies indicate the presence of A. hypogaea in the Huarmey Valley in Peru as long as 5,000 BP (Bonavia 1982).

Polyploids are sufficiently common among cultivated plants suggesting that they may have an advantage during artificial selection by man (Hilu 1993). This may be in part because of greater vigor due to 1) heterosis, and 2) increased size of harvested organs (Gepts 2003). In addition, events caused by polyploidy such as 3) changes in gene expression through the increased possibilities offered by: higher gene dosage, differential silencing, creation of new diversity through “genomic shock”; activation of transposons, and 4) relaxed selection on duplicated genes granting the acquisition of new gene functions are also likely to be important factors in allowing greater adaptability to cultivation (Soltis and Soltis 1995; Wessler and Carrington 2005). Whatever are the exact molecular mechanisms involved, it is remarkable that allotetraploid peanut, which has a very narrow genetic base, was transformed by domestication into one of the world's most important crops: completely distinct in plant architecture, seed size and pod form from its wild ancestors. In contrast, the much more genetically diverse diploid species, which have been cultivated for at least the same amount of time, only gave rise to a couple of proto-domesticate species cultivated on a very limited scale by indigenous people to this day (JFM Valls, personal communication; Freitas 2004).

2. 
   Some Aspects of the Genetic Behavior of Wild and Cultivated Peanuts
   

The center of diversity of the genus Arachis is in the Cerrado biome, a savannah-like vegetation that experiences highly seasonal rainfall with very distinct wet and dry seasons (Krapovickas and Gregory 1994; Valls and Simpson 2005). Perhaps as an evolved adaptation to these climatic conditions, all peanut species have an unusual reproductive biology. They bear their fruits under the ground, or in other words, they are geocarpic. The flowers are borne above ground, and, after fertilization the young fruit is pushed into the ground through an elongated portion of the fruit structure (“peg”) (Carla Pelegrin, personal communication) that has a specialized resistant tip. It penetrates the soil, a few centimetres for cultivated peanut, and much further for the wild species and the pod develops underground (Smith 1950). Deposited below the soil surface, peanut seeds are afforded protection from many pests and predators, favorable conditions for germination, and privileged access to soil moisture. However, a buried seed cannot be efficiently dispersed, and in natural conditions dispersal is mostly limited to the area covered by the maternal plant. More rarely, seeds may be deposited further afield by water-driven soil erosion, or animals (including man). In these cases, a single or very few seeds then found a new population, which results in natural populations that are “patches” with typically only tens to hundreds of individuals.

The combination of multiple recurrent severe genetic bottlenecks, small population sizes and a typically high rate of self-fertilization have provided the perfect conditions for genetic drift and the evolution of genetic isolation. In the classic Bateson–Dobzhansky–Muller model for the evolution of sexual incompatibility, diverging lineages evolve by mutations at different loci that are innocuous in their native genomic context, but interact negatively in hybrids (Bomblies and Weigel 2007). In the context of Arachis, its reproductive biology provides the perfect scenario for the fixation of weakly deleterious mutations that affect reproduction or indeed any other aspect of the plant’s biology. Once a mutation is fixed within a population, any compensatory mutation in the same or an interacting gene will be positively selected, thus driving forward species divergence and genetic isolation. These mechanisms may have caused the remarkable degree of sexual incompatibility observed between different collections classified as the same species of wild Arachis (Krapovikas and Gregory 1994). At a genomic level, we may expect the genomes of wild species to harbor the signatures of genetic bottlenecks, isolation, inbreeding and genetic drift.

A very severe genetic bottle was imposed at the origin of peanut, and with cultivation came different population dynamics. With man actively transporting seeds, populations became mobile, and would be likely to experience more genetic mixing. However, considering the low densities of itinerant farmers in prehistory, genetic bottlenecks could still easily have occurred. Also, a new genetic phenomenon may have accompanied domestication: the selective sweep. Here, through strong artificial selection, agronomically favorable alleles and their surrounding genomic regions rapidly spread through populations and are genetically fixed. Such sweeps leave characteristic signals in genome sequences, although they may be difficult to detect in a background of very low DNA polymorphism.

1. 
   The A and B Genomes of Peanut
   

Although peanut is an allotetraploid, chromosome pairing during meiosis is almost entirely bivalent (Smartt 1990). This presumably indicates that the A and B chromosomes have significantly diverged at the molecular level. Recently, the date of evolutionary divergence of the A and B genomes has been estimated for the first time. Sets of orthologous sequences were obtained from four intron regions, from A. duranensis and A. ipaënsis, Lotus, and both palaeopolyploid components of soybean. Calibrating a molecular clock of DNA sequence divergence using the known divergence dates of the two genomic components of soybean (13 Mya), and the Galegoid, Phaseoloid and Dalbergioid clades, the divergence of the Arachis A and B genomes was estimated at 3.5 Mya (Nielen et al. 2011). Although such estimations must always be used with caution, we can confidently say that A and B peanut genomes are much more similar than the soya subgenomes, and almost certainly more similar than the allotetraploid cotton subgenomes, which have an estimated divergence of 5-7 million years.

In terms of evolution, 3.5 million years is a relatively short time. However, it is ample time for very significant transposon activity. Indeed, most easily dated transposons in plant genomes are less than three million years old. Older elements tend to be degraded by mutation or eliminated by unequal crossing-over and illegitimate recombination (Vicient et al. 1999; Devos et al. 2002; Pereira 2004). A substantial divergence in the repetitive component of the two genome components of peanut is consistent with in situ hybridization experiments where chromosome spreads were probed with A. duranensis BAC clones (Fig. 1), (Guimarães et al 2008; Araujo et al. 2012), or with GISH using whole genomic DNA of its most probable ancestral diploids A. duranensis and A. ipaënsis (Seijo et al. 2007). The genomic probes do not hybridize exclusively, but predominantly to the chromosomes of their respective genome components. This shows that the repetitive components of the ancestral species diverged substantially during their separate evolutionary journeys traced since the time of their most recent common progenitor. Also, that any movement of repetitive DNA between the A and B genomes that occurred since the formation of the allotetraploid species was not sufficient to homogenize their repetitive DNA contents. Furthermore, the absence of any significant mosaic or chimeric hybridization patterns indicates that no large translocations between the A and B chromosomes have occurred since polyploidization. Indeed, the genome of peanut observed by GISH (Fig. 1a) is not distinguishable from the genome of a synthetic allotetraploid made from A. ipaënsis and A. duranensis (Fávero et al. 2006). Overall, it seems that there have been no obvious major structural changes in the diploid ancestral genomes following polyploidization.

The details of the GISH hybridization patterns could also be informative as to the distribution of repetitive DNA within the chromosomes. The strongest hybridization signals are at the interstitial chromosome regions. Hybridization is not detectable at the centromeres or the chromosome ends (Seijo et al. 2007; Nielen et al. 2010). Additionally, the smallest A chromosome pair, which has the most pronounced heterochromatic band, exhibits only very weak hybridization signals. From this, it is tempting to conclude that the repetitive DNA content of centromere and terminal regions of chromosomes are distinct from the interstitial regions and that the A chromosome pairs have distinct repetitive DNA profiles. However, it is also possible that the different hybridization patterns may, at least in part, be due to different states of chromatin condensation that influence the access of probes.

FISH patterns are also informative as to the distribution of genes and repetitive DNA within genome. Hybridization signals in A. hypogaea metaphase chromosomes using some gene-poor clones from A. duranensis bacterial artificial chromosome (BAC) library (A genome) as probes, were scatteredly distributed at the interstitial chromosome regions mainly in A chromosomes and eventually also in B (Fig. 1d), suggesting that the repetitive content of the A and B genomes is diverged. Patterns of hybridization such as intensity, diffused or spotted, varied according to the clone used as probe (Fig. 1b-d). Hybridization was generally at the pericentromere region but not at telomeres (Fig. 1 b-d). Preliminary sequence analysis indicated that most of the repetitive sequences present in these clones could be accounted for by multiple copies of just few long terminal repeat (LTR) retrotransposons and part of them. Additional analyses are underway to further explore genome structure that will help the evolutionary understanding and to delineate strategies to help the assemblage of full genome sequence.

2. 
   The Size and Repetitive Content of the Peanut Genome
   

Estimates of genome size vary depending on the methodology and standards used for its determination. The first estimate of genome size of peanut (Singh et al. 1996) was later re-evaluated substantially downwards to an equivalent of about 2.8 Gbp (Temsch and Greilhuber 2000, 2001). The size estimates for the most probable ancestral A and B genome donors, A. duranensis and A. ipaënsis, have been reported as being roughly similar to each other, which is consistent with observed chromosome sizes. The overall repetitive structure of the peanut genome has been studied by the kinetics of renaturation (Dhillon et al. 1980; Table 1). The high estimated percentage of repetitive DNA is within the range expected for genomes of this size.

Table 1 Repetitive structure of the peanut genome according to renaturation kinetics (data from Dhillon et al 1980).

% of Genome	Class	Average copy number
11.90%	High repeat	38,000
14.80%	Intermediate repeat	6,700
37.40%	Rarely repeated	200
36.00%	Single copy	1

Eukaryotic genomes harbor different types of repetitive DNA. Here we shall discuss ribosomal DNAs (rDNAs) and transposons.

2.2.1. rDNAs in Peanut and Its Diploid Ancestors

The most thorough study of the location and number of rDNAs was conducted by Seijo and collaborators (2004) using fluorescent in situ hybridization (FISH). The study showed, as previously mentioned, that the number, size, and distribution of rDNA clusters in A. hypogaea are virtually equivalent to the sum of those present in A. duranensis and A. ipaënsis. A single pair of 5S sites is present on each of the A and B chromosome complements, and two pairs of 18S-25S sites on the A chromosomes and three pairs on the B. The only exception to this equivalence is that in both of the diploid species, 18S-25S sites bear a thread-like constriction indicating intense transcriptional activity (forming the SAT chromosome; Fernandez and Krapovickas 1994). However, in the allotetraploid the constrictions are observed only on the A genome. This indicates that the transcriptional activity of the B genome rDNAs has been silenced, a common event in polyploids called nucleolar dominance (Cermeno et al. 1984; Preuss and Pikaard 2007).

2.2.2. Transposons and Their Evolution in the A and B Genomes

Transposons can be divided into two classes depending on whether their transposition intermediate is RNA (class 1, or retrotransposons) or DNA (class 2, or DNA transposons). Archetypal members of each group encode the protein products required for their transposition and are autonomous in function. However in a reductio ad absurdum of the rhyme:

“Big fleas have little fleas,

upon their backs to bite 'em,

and little fleas have lesser fleas,

and so on, ad infinitum..,”

even transposons are not free of parasites! In both classes of transposons there exist “parasitic” members with incomplete, degraded or completely absent coding regions. These transposons are non-autonomous, and depend on the proteins encoded by other elements for transposition. Plant genomes harbor a great diversity of transposons. However, two types, miniature inverted-repeat transposable elements (MITEs) and long terminal repeat (LTR) retrotransposons have made particularly notable contributions to plant genome organization and evolution (Feschotte et al. 2002).

MITEs are non-autonomous DNA transposons of less than 600 bp in length. In peanut, Patel and collaborators (Patel et al. 2004) reported that, following treatment with a chemical mutagen, a MITE insertion caused functional disruption of the fatty-acid desaturase-encoding gene ahFAD2B, one of the homeologous genes controlling the very important quality trait of high oleic / linoleic fatty acid ratio in peanut seeds. This MITE did not belong to the most common Tourist or Stowaway families but showed similarities to the Bigfoot family in Medicago (Charrier et al. 1999). Later, AhMITE1, a transposon with sequence similarities to the previously reported MITE, was observed to excise from a single locus in spontaneous and artificially induced mutants (Gowda et al. 2010, 2011).

Evidence of activity and a tendency to transpose into genes or their flanking regions (Feschotte et al. 2002) stimulated further interest in MITEs, and recently a large-scale analysis in peanut has been completed (Shirasawa et al. in press). Using enriched genomic libraries, 504 unique AhMITE1 sequences and their flanking genomic regions were obtained and shown to group into six families. Intriguingly, southern blots showed multiple AhMITE1 copies in the genomes of A. magna (a wild diploid B genome species very closely related to A. ipaënsis) and A. hypogaea, but not in the genome of A. duranensis, the most probable A genome donor to peanut. This suggests that AhMITE1 elements amplified in the B genome, but not in the A genome after their divergence about 3.5 Mya. Surveying of AhMITE1 insertion sites in cultivated varieties by PCR showed 13% polymorphism within a small sample of Virginia Runner type and 30% polymorphism between three Virginia cultivars and a Spanish type. This clearly indicates large-scale activity of AhMITE1 elements since the formation of the cultivated peanut and indicated the possibility that transposition events from the B to the A genome may have occurred in this tetraploid. The distribution of AhMITE1 markers in all the linkage groups of the most dense linkage map for peanut produced to date support that this migration has happened (Shirasawa et al. unpublished data). This conclusion is compatible with the apparent equivalence in GISH patterns of peanut and synthetic allotetraploid mentioned above, because MITEs are small and their movement would not be expected to significantly change genomewide chromosome hybridization patterns.

The insertion rate of AhMITE1 into BLASTX detectable genes (10.5%) is much more frequent than would be expected by chance, and indicates that this family of transposons is likely to have affected the expression of numerous genes since the formation of the tetraploid, and may have had an important role in the generation of present-day morphological diversity of cultivated peanuts (Shirasawa et al. 2012).

The first comprehensively characterized peanut retrotransposon was an autonomous Ty3-gypsy type element of about 11,200 bp named FIDEL (Nielen et al. 2010). FISH analysis of peanut, dot blots and BAC-end sequences from A. duranensis and A. ipaënsis indicate that this element is more frequent in the A than in the B genome, with copy numbers of about 3,000 and 820 per haploid genome respectively (0.7% of the tetraploid genome). Phylogenetic analysis of reverse transcriptase sequences showed distinct evolution of FIDEL in the A and B genomes and indicated that FIDEL most probably underwent two major events of transposition and multiplication in the A genome after its evolutionary divergence from the B genome.

In contrast to AhMITE1, FIDEL is less frequent near single copy genes, a tendency that was observed using paired sequences from BAC clones (Nielen et al. 2010). Interestingly, this tendency could not be demonstrated with resistance gene homologs (see more on the association of transposons and resistance gene homologs later in this chapter). On a chromosome scale, the distribution of FIDEL; strongest hybridization in the interstitial regions of chromosome arms, and absence of detectable signal from centromeres, telomeric regions and nucleolar organizer region, closely resemble that of whole genomic DNA probes. This indicates that FIDEL may be an important component of the divergence of the repetitive DNA of the A and B genomes (Nielen et al. 2010).

Whilst FIDEL is common in the peanut genome, most copies are likely to be in a mutated and/or epigenetic silenced state. This is suggested by stop codons in coding regions and by the Ts/Tv ratio of >1.5:1 in the LTR sequences (Nielen et al. 2010). However, searching expressed sequenced tag (EST) data does reveal some activity, and more intriguingly, transcription from FIDEL seems to be up-regulated under drought and disease stress (Brasileiro et al. 2012). Whether this evidence for activation indicates the generation of functional proteins and transposition remains to be answered. Whichever, FIDEL’s localization in euchromatic regions suggests that it may have modified the expression of other genes, through insertional inactivation, or through the promotor activity of its LTRs.

More recently, a 6,179 bp autonomous Ty1-copia retrotransposon from the Bianca lineage named Matita has been characterized in peanut (Nielen et al. 2011). Matita is much less abundant than FIDEL with an estimated 520 copies in the haploid cultivated peanut genome. Also, in contrast to FIDEL, Matita is mainly located on the distal regions of chromosome arms and is of approximately equal frequency on both A and B chromosomes. Furthermore, phylogenetic analysis and molecular dating of transposition events suggest that although Matita has been active since the divergence of the A and B genomes, it underwent its last major burst of transposition activity at around the same time as the evolutionary divergence of peanut’s diploid ancestors. By probing BAC libraries it was shown that Matita is also not randomly distributed in the genome but exhibits a significant tendency of being more abundant near resistance gene homologs than near single copy genes.

These studies have given a glimpse of the importance of transposons in the evolution of the peanut genome, in terms of their influence on gene expression and on genome structure. It is a common theme in plant genome structure that a relatively few transposon species are present at high number, and many more are present in low numbers. Peanut seems to follow this pattern, as the sequence analysis of 12 A-genome BACs, spanning about 1.25 Mb has recently shown (Araujo et al. 2012). Within these BAC sequences, most of the repetitive sequences could be accounted for by multiple copies of just seven LTR retrotransposons, their solo LTRs and remnants that are under characterization by the authors (Araujo et al. unpublished data). Interestingly, only three of these elements were autonomous and four were non-autonomous, with one of the non-autonomous elements having FIDEL-like LTRs. Most of the datable transpositions were less than three Mya, indicating that much of the divergence of the A and B gene-space may be accounted by the activity of a few species of LTR retrotransposons. These elements are frequent in gene-space, but may be even more so in regions of the genome with few or no genes (see the example in Fig. 2).

3. 
   Comparison of the A and B Genomes in Low Copy Regions of the Genome
   

Whilst the consideration of repetitive DNA in the above section emphasizes the differences between the A and B genomes, comparisons of genetic maps (see later in this chapter) and lower copy DNA emphasize the similarities. At the moment there is little data, but sequence identity is on average 94% between two 45 kb A and B homeologous genome sequences characterized by the authors (Bertioli et al. unpublished data). This degree of sequence similarity in noncoding regions is compatible with the estimated date of evolutionary divergence (3.5 Mya) and a substitution rate of 1.3 x 10^-8 per site per year (Ma et al. 2004). This is likely to be representative of many regions of the genome. In coding regions that are under conservative selection, we may expect DNA sequence identity to be even higher. However, these regions of high identity will be “broken” by fast evolving repetitive DNA. We may expect that many of these breaks can be accounted by the activity of a rather few species of transposons (Araujo et al. 2012).

We have already seen that the distributions of AhMITE1, FIDEL and Matita are not spatially random. Over evolutionary time, nonrandom patterns of transposon activity and elimination from the genome create a genomic landscape with identifiable broad features. In the next section we will see what light genetic mapping has shed on this.

3.