Phylogeny and Systematics of Crop Plants

Searching for Wild Relatives and Gene Donors

Modern crop plants possess a narrow genetic base. During domestication and breeding, many alleles were lost, and genetic diversity decreased (Connor et al. 2011). Wild relatives preserve alleles that can provide resistance to diseases, pests, drought, salinity, and other stresses. The search for such donors begins with phylogenetic analysis: we determine which species are the closest relatives of the crop, and then evaluate their economically valuable traits.

Example: rice Oryza sativa is sensitive to salinity, whereas its wild relative Oryza coarctata is able to grow in coastal conditions flooded with seawater. Studying such relatives makes it possible to identify genes and alleles lost during domestication (Chen et al. 2024). Similarly, in tomato Solanum lycopersicum, breeding for large fruit led to the loss of alleles associated with salt tolerance, for example in the genes SlHAK20 and SlOSO1 (Chen et al. 2024). Phylogeny indicates where to look for these alleles.

A classic example of the successful use of a wild relative is tomato. The species Solanum chmielewskii, a close relative of the cultivated tomato, was used in crosses to improve fruit flavor. This became possible precisely because phylogenetic relationships showed the closeness of these species (Judd et al. 2018).

Predicting Crossability and Introgression

Phylogenetic proximity is one of the key factors determining the success of hybridization. The closer the species, the higher the probability of obtaining fertile hybrids and gene introgression. In breeding, this is expressed in the concept of gene pools: the primary gene pool (GP1) includes intercrossing forms, the secondary (GP2) includes species with which hybridization is possible with difficulties, and the tertiary (GP3) includes species requiring special methods (embryo rescue, bridge crosses). These concepts rely on phylogenetic data.

In addition, reproductive biology itself is evolutionarily determined. Many crops possess self-incompatibility (SI), which affects fruit and seed set. A distinction is made between gametophytic and sporophytic self-incompatibility, heteromorphic and late self-incompatibility (Chabert & Mallinger 2025). Understanding these mechanisms is necessary for planning planting schemes, selecting pollinators, and choosing pollinator varieties. For example, for cross-pollinated crops with self-incompatibility, compatible pollinator varieties must be planted. In addition, there is the phenomenon of xenia—the influence of the pollen donor on seed and fruit quality—which must also be taken into account when selecting varieties (Chabert & Mallinger 2025).

Polyploidy is another factor closely linked to phylogeny and breeding. Autopolyploids and allopolyploids arise as a result of genome doubling or hybridization. Polyploid complexes often include both cultivated and wild forms, and understanding their origin helps predict crossability and trait inheritance. For example, many crop plants are allopolyploids, and their phylogenetic history includes several parental species.

Reconstructing the Evolution of Economically Valuable Traits

Phylogeny makes it possible to reconstruct when and in what sequence economically valuable traits arose. For example, loss of grain shattering in cereals, increased fruit size in tomato, loss of bitterness in cucumber, and formation of non-dehiscent fruits are all results of changes in developmental networks (Ambrose & Ferrandiz 2013). By comparing traits in cultivated and wild forms, we can determine which changes are ancient and which are recent.

The concept of the “evolutionary depth” of a trait is important for breeding. If a trait is controlled by ancient, conserved regulatory cascades, its transfer may be more difficult than the transfer of a recently arisen trait. For example, in artichoke and cardoon, domestication followed different paths: in artichoke, large flower heads were selected; in cardoon, fleshy petioles. These changes involve different sets of genes and different developmental stages (Lanteri & Portis 2008). The phylogenetic context helps understand which traits are the result of parallel evolution and which are homologous.

Managing Weedy and Feral Forms

Weedy plants and feral relatives of crops often originate from the same species as crop plants. For example, red rice in cultivated rice fields and shattercane in sorghum are forms that returned to a “wild” phenotype as a result of natural selection (Connor et al. 2011). Knowledge of phylogeny and genetic relationships makes it possible to predict the appearance of such forms, assess the risk of cross-pollination, and develop control measures. In addition, understanding the evolution of herbicide resistance helps plan crop rotations and use different herbicide modes of action (Connor et al. 2011).

Conservation and Use of Genetic Resources

Gene banks and germplasm collections play a key role in conserving the diversity of crops and their wild relatives. However, collections must be representative and manageable. Phylogenetic analysis helps identify core collections—minimal sets of accessions that cover maximum genetic diversity (Connor et al. 2011). For example, for artichoke, core collections were created on the basis of AFLP markers, which made it possible to organize work with germplasm (Lanteri & Portis 2008). Similarly, for many crops, in situ and ex situ conservation strategies are being developed based on phylogenetic data.

Breeding for Resistance to Climatic Stresses

Climate change makes the search for sources of drought tolerance and salt tolerance especially urgent. Xerophytes and salt-tolerant plants possess a number of adaptations: effective sequestration of Na⁺ in vacuoles, operation of NHX transporters and H⁺-pyrophosphatases, altered structure of ABA receptors, fewer stomata, but faster stomatal response (Chen et al. 2024). Phylogenetic analysis shows which of these traits are ancient and which arose independently in different lineages. This helps purposefully search for donors among wild relatives of crops.

For example, in tomato, alleles associated with salt tolerance were lost during domestication. Comparison with wild relatives makes it possible to identify these alleles and return them to cultivated varieties (Chen et al. 2024). Similarly, for rice, the wild species Oryza coarctata is a potential donor of salt-tolerance genes.

Understanding Reproductive Biology and Pollination

Phylogeny and systematics help understand the reproductive strategies of crops. For example, self-incompatibility (SI) and early-acting inbreeding depression (EID) can limit seed set under self-pollination. For crops with EID, such as blueberry Vaccinium spp., it is important to ensure cross-pollination, since self-pollination leads to embryo abortion (Chabert & Mallinger 2025). Xenia—the influence of the pollen donor on fruit and seed quality—may also be important for planning varietal mixtures (Chabert & Mallinger 2025). All of this is directly related to phylogenetic relationships and reproductive compatibility.

Tomato

Tomato Solanum lycopersicum has many wild relatives in the genus Solanum. One of them, S. chmielewskii, was used to improve fruit flavor. Charles Rick crossed the cultivated tomato with this wild species and obtained lines with improved flavor (Judd et al. 2018). This example shows how phylogenetic knowledge makes it possible to find valuable alleles in wild relatives.

In addition, studies of tomato domestication have shown that breeding for large fruit led to the loss of salt-tolerance alleles. The genes SlHAK20 and SlOSO1 are examples of such lost adaptations (Chen et al. 2024). Returning these alleles to cultivated varieties is a task of modern breeding based on phylogenetic comparison.

Rice

Rice Oryza sativa is one of the most important food crops. Its wild relative Oryza coarctata possesses exceptional salt tolerance and is able to grow under conditions of daily seawater flooding. Studying this species makes it possible to identify genes responsible for salt tolerance and transfer them into cultivated rice (Chen et al. 2024). Phylogenetic analysis shows that O. coarctata is a sufficiently close relative for hybridization to be possible, although it requires special methods.

Artichoke and Cardoon

Artichoke (Cynara cardunculus var. scolymus) and cardoon (C. cardunculus var. altilis) originate from wild cardoon (C. cardunculus var. sylvestris). Domestication proceeded independently: artichoke was selected for large flower heads, cardoon for fleshy petioles. Genetic studies have shown that these two forms are significantly differentiated, while a high level of diversity is preserved within cultivated forms (Lanteri & Portis 2008). Phylogenetic analysis made it possible to create a core collection and determine strategies for conserving germplasm. This example shows how phylogeny helps understand the origin of crops and organize work with genetic resources.

Wheat and Other Cereals

Wheat, rice, maize, and barley are major food crops. Their breeding has led to a significant increase in yield, but also to a narrowing of the genetic base. Comparison of historical varieties shows that genetic progress in yield is associated mainly with changes in harvest index and disease resistance, rather than with increased photosynthetic capacity (Connor et al. 2011). Understanding phylogeny and genetic resources makes it possible to purposefully search for new alleles for further improvement.

Which Genomes Are Used in Plant Phylogeny

Plants have three genomes: nuclear, chloroplast, and mitochondrial. Chloroplast and mitochondrial genomes are usually inherited uniparentally, most often maternally, whereas the nuclear genome is biparental (Pandey 2021). This determines their advantages and limitations.

Chloroplast genome (cpDNA) is the most popular source of markers for plant phylogeny. It has a small length (usually 120–160 thousand bp), contains single- and two-copy regions, inverted repeats, and is inherited predominantly maternally. Key advantages: small size, uniparental inheritance, conserved structure, relatively high rate of nucleotide substitutions compared with the mitochondrial genome, and the possibility of amplification with standard primers (Pandey 2021). This is precisely why cpDNA became the basis for phylogenetic reconstructions at the level of genera, families, and above.

Among the most frequently used cpDNA loci:

• rbcL — the gene for the large subunit of RUBISCO, slowly evolving, about 1430 bp long, located in the large single-copy region. Good for reconstructions at the genus level and above (Pandey 2021).
• matK — the maturase K gene, located in the intron of the trnK gene, about 1550 bp long, evolves rapidly, has a low transition-to-transversion ratio, and is widely used for distinguishing species and genera (Pandey 2021).
• ndhF — the NADH dehydrogenase F gene, located in the small single-copy region, about 2235 bp long, highly variable, used at family, subfamily, tribal, and genus levels (Pandey 2021).
• atpB — the beta-subunit gene of ATP synthase, about 1497 bp long, close to rbcL in evolutionary rate, contains no introns, and aligns easily (Pandey 2021).
• rpl16 — a noncoding intron, about 1059 bp, with frequent insertions and deletions, useful for phylogeny at lower taxonomic levels (Pandey 2021).
• trnH–psbA — an intergenic spacer, one of the most variable in angiosperms, from 296 to 1120 bp long, used for species delimitation, but limited by its short length and possible inversions (Pandey 2021).
• trnL–trnF — the intron of the trnL (UAA) gene and the intergenic spacer between trnL and trnF, highly variable, convenient for phylogeny at the genus and species levels, sometimes up to tribes (Pandey 2021).

Mitochondrial genome (mtDNA) in plants is larger and more variable in size (from 200 to 2500 thousand bp), characterized by high intramolecular recombination and a low substitution rate. Therefore, it is used less often in plant systematics, although individual loci (for example, cox1, atpA) may be useful (Pandey 2021).

Nuclear genome is the largest and most complex. For plant phylogeny, ribosomal genes are widely used: 18S, 26S, 5.8S, as well as internal transcribed spacers ITS1 and ITS2. The ITS region (ITS1 + 5.8S + ITS2) has a length of 600–700 bp, is easily amplified with universal primers, has a high substitution rate, and is often used for phylogeny at the species and genus levels. It is called the “workhorse” of plant molecular systematics (Pandey 2021). The external transcribed spacer (ETS) is also used, but its use is limited because of the variable nontranscribed spacer (NTS) at the 5′ end (Pandey 2021).

Modern phylogenetic studies increasingly use low-copy nuclear genes and whole genomes. For example, APG IV relies on analysis of many genes, including nuclear, chloroplast, and mitochondrial (APG IV 2016). This makes it possible to overcome the limitations of individual markers and obtain a more reliable reconstruction.

The Problem of Choosing Markers for Crop Plants

When choosing markers for the phylogeny of a specific crop, several factors must be taken into account.

First, the level of taxonomic differentiation. For closely related species and varieties, rapidly evolving markers are needed (ITS, trnH–psbA, microsatellites). For deeper nodes, conservative markers are needed (rbcL, 18S).

Second, hybridization and polyploidy. Many crop plants arose as a result of hybridization and chromosome doubling. In such cases, nuclear markers may give “reticulate” relationships, while chloroplast markers reflect only the maternal line. Therefore, for allopolyploids, several nuclear loci must be analyzed to identify parental genomes (Judd et al. 2018; APG IV 2016).

Third, intraspecific variability. Crop plants are often represented by many varieties, landraces, and wild forms. The sample should include as many accessions as possible in order to cover genetic diversity. Otherwise, the tree may reflect not evolution but sampling artifacts (Pandey 2021).

Fourth, the availability of reference sequences. For many crops, complete genomes and databases, such as GenBank, already exist. This accelerates marker selection and comparison (Pandey 2021).

Methods of Constructing Phylogenetic Trees

After obtaining aligned sequences, it is necessary to choose a method for tree construction. In molecular systematics, two main groups of methods are used (Pandey 2021):

1. Distance methods — algorithmic, fast, and produce one tree. These include:

• UPGMA (Unweighted Pair-Group Method with Arithmetic Averages) — a clustering method assuming ultrametricity, that is, equal distance of all taxa from the root. It constructs a phenogram, not a phylogram. It is good for genetic frequencies, but rarely suitable for molecular phylogeny, where evolutionary rates may differ (Pandey 2021).
• Neighbor-Joining (NJ) — a minimum-evolution method, constructs an unrooted tree, works quickly with large samples, and does not evaluate tree quality. It is widely used for preliminary analysis (Pandey 2021).

2. Tree-search methods — evaluate many trees and choose the best according to a certain criterion:

• Maximum Parsimony (MP) — chooses the tree with the minimum number of evolutionary changes. Good for morphological and molecular data, but may suffer from long-branch attraction (Pandey 2021).
• Maximum Likelihood (ML) — chooses the tree that most likely explains the data under a given model of evolution. It takes substitution frequencies into account, requires more computation, but gives more reliable estimates (Pandey 2021).
• Bayesian analysis — estimates the posterior distribution of trees and uses MCMC. It makes it possible to obtain posterior probabilities for nodes and estimate divergence times (Pandey 2021).

To assess node stability, bootstrap and posterior probabilities are used. Bootstrap values above 95% are considered well supported, 70–90% moderately, and below 50% unstable (Pandey 2021).

It is important to remember that a gene tree does not always coincide with a species tree. Genes may have their own history, especially under hybridization, incomplete lineage sorting, and horizontal transfer. Therefore, reliable reconstruction of a species tree requires multigene data and coalescent analysis methods (Judd et al. 2018; Pandey 2021).

Basic Elements of a Phylogenetic Tree

A phylogenetic tree consists of nodes and branches. Terminal nodes correspond to studied taxa; internal nodes correspond to hypothetical ancestral forms. The branch between two nodes is called an internode. The root is the node from which the entire tree originates; it determines the direction of evolution. An unrooted tree has no root and does not allow determination of trait polarity (Judd et al. 2018).

A clade is a group including an ancestor and all its descendants, that is, a monophyletic group. A paraphyletic group includes an ancestor but not all descendants. A polyphyletic group includes descendants of different ancestors. In phylogenetic systematics, only monophyletic groups are recognized (Judd et al. 2018; APG IV 2016).

Sister groups are two clades originating from one node. They are closest to each other. For example, if in a tree crops A and B are sister groups, then they are closer to each other than to C (Judd et al. 2018).

What We Look for on a Crop Tree

To read the phylogeny of a specific crop, it is necessary to:

1. Find the position of the cultivated species on the tree.

2. Determine its sister groups. These are the closest wild relatives.

3. Assess which taxa belong to the same clade as the crop. This is its gene pool.

4. Determine where the wild relatives (crop wild relatives, CWR) are located and how close they are.

5. Map geographic data onto the tree to infer centers of origin.

6. Map economically valuable traits onto the tree and reconstruct their evolution.

Determining Centers of Origin

Centers of origin of crop plants can be reconstructed by superimposing data on the modern distribution of wild relatives and cultivated forms onto the phylogenetic tree. If the sister groups of a crop are concentrated in a certain region, this suggests that domestication began precisely there. For example, for artichoke and cardoon, the wild ancestor Cynara cardunculus var. sylvestris is distributed in the Mediterranean, indicating this region as the center of origin (Lanteri & Portis 2008). For rice, wild Oryza species occur in Southeast Asia, which corresponds to the center of origin of the crop (Chen et al. 2024).

However, it is important to take into account that centers of origin may be multiple, and cultivated forms may have arisen repeatedly in different places. Phylogeny helps distinguish monophyletic origin from several independent domestications.

The Concept of Crop Wild Relatives (CWR)

Wild relatives of crops are species genetically close to crop plants that can be used in breeding as sources of valuable alleles. They often grow in natural habitats and preserve adaptations lost during domestication (Connor et al. 2011). The search for CWR begins with phylogenetic analysis: the sister groups of the crop are the first candidates.

Gene Pools GP1–GP3

The classical scheme of Harlan and de Wet divides the genetic resources of a crop into three pools:

• GP1 (primary gene pool) — forms that cross easily with the crop and produce fertile offspring. These include cultivated varieties, landraces, and some wild forms of the same species.
• GP2 (secondary gene pool) — species with which crossing is possible, but hybrids may be sterile or require special methods. For example, for wheat these are certain Aegilops species.
• GP3 (tertiary gene pool) — species with which crossing is possible only using embryo rescue, bridge crosses, or genetic engineering. For example, for tomato these are certain wild Solanum species (Judd et al. 2018; Connor et al. 2011).

Phylogenetic proximity often, although not always, correlates with belonging to one or another pool. The closer a species is on the tree, the higher the probability of successful hybridization.

Reproductive Compatibility and Self-Incompatibility

When planning crosses, reproductive biology must be taken into account. Many crops possess self-incompatibility (SI), which prevents self-pollination. A distinction is made between gametophytic (GSI), sporophytic (SSI), heteromorphic (HSI), late (LSI), and diallelic (DSI) self-incompatibility (Chabert & Mallinger 2025). For example, in apple and pear, GSI is widespread, in which pollen with the same S-allele does not germinate. To obtain a harvest, compatible pollinator varieties must be planted.

In addition, in self-compatible crops, early-acting inbreeding depression (EID) may occur, in which self-pollination leads to embryo abortion. This is characteristic of blueberry Vaccinium spp. and some other crops (Chabert & Mallinger 2025). In such cases, cross-pollination must be ensured.

There is also xenia—the influence of the pollen donor on seed and fruit quality. This is important to take into account when selecting pollinator varieties (Chabert & Mallinger 2025).

Polyploid Complexes

Polyploidy is a common phenomenon among crop plants. Autopolyploids arise through doubling of the genome of one species; allopolyploids arise through hybridization of different species followed by chromosome doubling. Polyploid complexes often include both cultivated and wild forms. For example, wheat Triticum aestivum is an allohexaploid, and potato Solanum tuberosum is an autopolyploid (Judd et al. 2018; Connor et al. 2011).

Phylogenetic analysis of polyploids requires a special approach: parental genomes must be identified, which is possible using several nuclear markers. Chloroplast markers will show only the maternal line.

Reconstruction Methods

After constructing a phylogenetic tree, the evolution of traits can be reconstructed. For this, methods of parsimony ancestral state reconstruction, maximum likelihood, and Bayesian are used. They make it possible to determine what state of a trait the ancestor had and at which nodes changes occurred (Judd et al. 2018).

Examples of Economically Valuable Traits

Large fruits. In tomato Solanum lycopersicum, increasing fruit size was one of the key directions of domestication. Comparison with wild relatives shows that large-fruitedness is associated with changes in regulatory networks controlling cell division and expansion (Ambrose & Ferrandiz 2013). Phylogenetic analysis makes it possible to determine which alleles were lost and which can be returned.

Non-dehiscent fruits. In many crops, selection favored loss of seed shattering. For example, in cereals, loss of shattering is associated with changes in genes controlling formation of the abscission layer. In artichoke and cardoon, selection favored different traits: in artichoke, large flower heads; in cardoon, fleshy petioles. Genetic studies have shown that the spiny trait in artichoke is controlled by the Sp locus, flower head color by the P and U loci, and density by the C and H loci (Lanteri & Portis 2008). These data make it possible to conduct breeding purposefully.

Loss of bitterness. In cucumber Cucumis sativus, wild forms contain bitter cucurbitacins, while cultivated forms do not. Loss of bitterness is associated with mutations in genes of cucurbitacin biosynthesis. Phylogenetic analysis shows that this occurred independently in different lineages (Judd et al. 2018).

Fleshy fruits. Fleshy fruits arose repeatedly in different families. For example, in Solanaceae, Rosaceae, and Vitaceae. This is an example of parallelisms—independent evolution of similar adaptations. Phylogeny helps distinguish homology from analogy.

Stress resistance. In xerophytes and salt-tolerant plants, common adaptations have been identified: an increase in the number of copies of NHX and H⁺-PPase genes, altered structure of ABA receptors PYR/PYL, reduced stomatal density, but increased speed of their response (Chen et al. 2024). Phylogenetic analysis shows which of these traits are ancient and which arose independently in different lineages. This helps search for donors among wild relatives.

Homology, Analogy, and Parallelisms

It is important to distinguish:

• Homology — similarity due to common origin. For example, all flowers of angiosperms are homologous as structures, although they may differ greatly.
• Analogy — independent origin of similar traits. For example, wings in birds and insects.
• Parallelisms — independent evolution of similar adaptations in closely related species. For example, fleshy fruits in different genera of Solanaceae.

For breeding, this is fundamental: if a trait is homologous, it can be transferred between closely related species; if analogous, different approaches may be required. Phylogenetic analysis combined with developmental studies (evo-devo) makes it possible to distinguish these cases (Ambrose & Ferrandiz 2013; Judd et al. 2018).

What CWR Are and Why They Are Important

Crop wild relatives (CWR) are species genetically close to crop plants that can be used in breeding as sources of valuable alleles. They grow in natural habitats and preserve adaptations lost during domestication (Connor et al. 2011). CWR are part of overall genetic diversity, and their conservation and study are one of the key directions of modern agronomy.

During domestication and breeding, the genetic diversity of crop plants decreased significantly. This process is called the “domestication bottleneck.” Humans selected plants according to a limited set of traits: large fruits, non-dehiscent seeds, absence of bitterness, synchronous ripening. As a result, many alleles associated with resistance to diseases, pests, drought, salinity, and other stresses were lost (Connor et al. 2011; Chen et al. 2024).

CWR preserve these alleles. For example, in rice Oryza sativa, salt tolerance was lost during domestication, whereas its wild relative Oryza coarctata is able to grow in coastal conditions flooded with seawater. Studying such relatives makes it possible to identify genes and alleles lost during domestication (Chen et al. 2024). Similarly, in tomato Solanum lycopersicum, breeding for large fruit led to the loss of salt-tolerance alleles associated with the genes SlHAK20 and SlOSO1 (Chen et al. 2024). Phylogenetic analysis indicates where to look for these alleles.

Classification of CWR by Degree of Kinship

CWR can be classified by the degree of genetic proximity to the crop plant. This classification is directly related to the concept of gene pools (GP1–GP3), which we will consider below. The closer a species is on the phylogenetic tree, the higher the probability of successful hybridization and gene introgression.

The sister groups of a crop are the first candidates for the role of donors. They are the closest and, consequently, the most compatible. However, the sister group does not always possess the needed traits. Sometimes valuable alleles are found in more distant relatives, which requires special methods of transfer.

Examples of CWR Use in Breeding

Tomato. The species Solanum chmielewskii, a close relative of the cultivated tomato, was used to improve fruit flavor. Charles Rick crossed the cultivated tomato with this wild species and obtained lines with improved flavor (Judd et al. 2018). This example shows how phylogenetic knowledge makes it possible to find valuable alleles in wild relatives.

Wheat. For wheat Triticum aestivum, the most important source of disease-resistance genes are species of Aegilops. For example, rust-resistance genes were transferred from Aegilops into cultivated wheat. These species belong to the secondary gene pool (GP2) and require special hybridization methods.

Rice. The wild species Oryza coarctata is a potential donor of salt-tolerance genes. Studying this species makes it possible to identify genes responsible for salt tolerance and transfer them into cultivated rice (Chen et al. 2024).

Artichoke and cardoon. Wild cardoon Cynara cardunculus var. sylvestris is the ancestor of both artichoke and cardoon. It preserves genetic diversity that can be used to improve cultivated forms. For example, resistance to diseases, earliness, dwarfness, and other traits can be transferred from wild forms (Lanteri & Portis 2008).

Problems of CWR Conservation

CWR are threatened with extinction due to habitat destruction, climate change, and agricultural activity. Conservation of CWR requires both in situ (in natural habitats) and ex situ (in gene banks) measures. Gene banks, such as collections of international research centers, play a key role in conserving CWR. However, collections must be representative and manageable. Phylogenetic analysis helps identify core collections—minimal sets of accessions that cover maximum genetic diversity (Connor et al. 2011).

Hybridization in Nature

Hybridization is crossing between different species or genetically distinct populations. In nature, hybridization occurs when closely related species come into contact, especially in zones of sympatry. Hybrids may be fertile or sterile, depending on the degree of genetic divergence of the parental species.

Hybridization plays an important role in plant evolution. It can lead to the emergence of new species (hybrid speciation), especially under polyploidy. For example, many allopolyploids arose as a result of hybridization between different species followed by chromosome doubling (Judd et al. 2018).

Gene Introgression

Introgression is the transfer of genes from one species to another through hybridization and subsequent backcrosses with one of the parental species. Introgression can occur in nature and in breeding. In nature, introgression can enrich the gene pool of a species with new alleles. In breeding, introgression is used to transfer valuable alleles from wild relatives into cultivated varieties.

Example: in artichoke and cardoon, wild cardoon C. cardunculus var. sylvestris crosses fully with cultivated forms, and their F₁ hybrids are fertile. This makes it possible to use wild forms as sources of genes for improving crops (Lanteri & Portis 2008).

Hybridization and Breeding

In breeding, hybridization is the main method of creating new varieties. It makes it possible to combine valuable traits from different parents. However, hybridization between distant species may be difficult due to reproductive barriers. To overcome these barriers, special methods are used: embryo rescue, bridge crosses, in vitro culture, and polyploidization.

Embryo rescue is used when the hybrid embryo aborts at early stages of development. The embryo is extracted and cultured in vitro. This method is widely used to obtain hybrids between crop plants and their distant wild relatives.

Bridge crosses are used when a direct hybrid between two species is impossible. First, a hybrid is obtained between one of the parents and a third species (the bridge), and then it is crossed with the other parent. This makes it possible to overcome reproductive barriers.

Polyploidization is used to restore fertility of sterile hybrids. Chromosome doubling leads to the formation of allopolyploids, which may be fertile because each chromosome has a homolog for conjugation in meiosis.

Hybridization and Weedy Forms

Hybridization between crop plants and their wild relatives can lead to the emergence of weedy forms. For example, red rice in cultivated rice fields is a form that arose as a result of hybridization and natural selection. Such forms can compete with the crop and reduce yield. Understanding phylogenetic relationships and hybridization processes helps predict the appearance of weedy forms and develop control measures (Connor et al. 2011).

Types of Polyploidy

Polyploidy is the presence of three or more complete chromosome sets in a cell. Two main types of polyploidy are distinguished:

• Autopolyploidy — doubling of the genome of one species. For example, Solanum tuberosum (potato) is an autopolyploid. Autopolyploids often arise through fusion of unreduced gametes (2n) with normal ones (n) or through somatic chromosome doubling.
• Allopolyploidy — hybridization of different species followed by chromosome doubling. For example, Triticum aestivum (bread wheat) is an allohexaploid, Gossypium barbadense (cotton) is an allotetraploid, Nicotiana tabacum (tobacco) is an allotetraploid (Judd et al. 2018; Connor et al. 2011).

Autopolyploids

Autopolyploids contain three or more homologous chromosome sets. In meiosis, this can lead to the formation of multivalents—conjugation of more than two chromosomes. Multivalents can lead to unequal chromosome segregation and, consequently, reduced fertility. However, autopolyploids can be fertile if conjugation occurs predominantly as bivalents.

Autopolyploidy is widespread among crop plants. For example, Tolmeia menziesii (autotetraploid, 2n = 28), Lythrum salicaria (autotetraploid, 2n = 60), Phleum pratense (autohexaploid, 2n = 42) (Judd et al. 2018). In alfalfa Medicago sativa, autopolyploidy also occurs.

Autopolyploids can arise spontaneously in nature and in cultivation. They often possess increased cell and organ size, which may be economically valuable. However, they can also suffer from inbreeding depression and genetic instability.

Allopolyploids

Allopolyploids arise as a result of hybridization of two different species followed by chromosome doubling. They contain complete genomes of both parental species. In meiosis, chromosomes of each genome conjugate with homologs from the same genome, which ensures fertility.

Allopolyploidy is a major mechanism of speciation in plants. Many crop plants are allopolyploids:

• Triticum aestivum (bread wheat) — allohexaploid (2n = 6x = 42). Its genomes originate from three parental species: T. urartu (A genome), Aegilops speltoides (B genome), and Ae. tauschii (D genome).
• Gossypium barbadense (cotton) — allotetraploid (2n = 52).
• Nicotiana tabacum (tobacco) — allotetraploid (2n = 48).
• Vaccinium corymbosum (highbush blueberry) — allotetraploid (2n = 48) (Judd et al. 2018).

Allopolyploids often possess hybrid vigor (heterosis) and can combine valuable traits of both parents. They may also be more resistant to stresses and diseases.

Polyploid Complexes

Polyploid complexes are groups of closely related species with different levels of ploidy. Such complexes often include diploid, tetraploid, hexaploid, and other forms. Polyploid complexes are widespread among crop plants and their wild relatives.

Example: in blueberry Vaccinium, there are diploid, tetraploid, and hexaploid species. Inbreeding depression in polyploid forms may be lower than in diploids, which is associated with the buffering effect of polyploidy (Chabert & Mallinger 2025). However, this effect decreases over time after polyploidization.

Example: in artichoke and cardoon Cynara cardunculus, polyploidy also plays a role. Wild cardoon, artichoke, and cultivated cardoon cross fully, and their F₁ hybrids are fertile. This makes it possible to use wild forms as sources of genes (Lanteri & Portis 2008).

Polyploidy and Breeding

Polyploidy is widely used in breeding. Artificial polyploidization is achieved by treatment with colchicine, which blocks spindle formation and leads to chromosome doubling. This makes it possible to:

• restore fertility of sterile hybrids;
• increase organ size;
• increase resistance to stresses;
• create new forms with valuable traits.

For example, triploid forms of some crops (banana, watermelon) possess increased productivity but are sterile. They are propagated vegetatively.

However, polyploidy can also have negative consequences: genetic instability, meiotic anomalies, reduced fertility, increased genome size, which can slow growth and development.

Determining Ploidy

To determine ploidy, cytological methods are used (counting chromosomes in metaphase of mitosis or meiosis) and flow cytometry. Flow cytometry makes it possible to quickly assess the amount of DNA in the nucleus and, consequently, the level of ploidy. This method is widely used for screening large samples (Judd et al. 2018). For example, in hawthorn Crataegus, flow cytometry made it possible to identify diploid, triploid, and tetraploid forms (Judd et al. 2018).

The Concept of Gene Pools

The concept of gene pools was proposed by Harlan and de Wet to describe the genetic resources of crop plants. It divides all forms related to a crop into three pools depending on the ease of crossing and fertility of hybrids.

• GP1 (primary gene pool) — forms that cross easily with the crop and produce fertile offspring. These include cultivated varieties, landraces, and some wild forms of the same species. Crossing within GP1 does not require special methods.
• GP2 (secondary gene pool) — species with which crossing is possible, but hybrids may be sterile or require special methods (embryo rescue, bridge crosses). For example, for wheat these are certain Aegilops species.
• GP3 (tertiary gene pool) — species with which crossing is possible only using embryo rescue, bridge crosses, or genetic engineering. Hybrids are often sterile. For example, for tomato these are certain wild Solanum species (Judd et al. 2018; Connor et al. 2011).

Examples of Gene Pools

Wheat. GP1 includes cultivated varieties Triticum aestivum and T. durum, as well as wild forms T. urartu and Aegilops tauschii (donors of the A and D genomes). GP2 includes other Aegilops and Triticum species that can cross with wheat, but with difficulties. GP3 includes more distant cereal species, crossing with which requires special methods.

Rice. GP1 includes cultivated rice Oryza sativa and its wild forms O. rufipogon and O. nivara. GP2 includes other Oryza species, such as O. officinalis and O. australiensis. GP3 includes even more distant species, such as O. coarctata (donor of salt tolerance) (Chen et al. 2024).

Tomato. GP1 includes cultivated tomato Solanum lycopersicum and its close wild relatives, such as S. pimpinellifolium. GP2 includes species S. peruvianum and S. chilense, crossing with which requires embryo rescue. GP3 includes more distant Solanum species.

Artichoke and cardoon. GP1 includes cultivated artichoke, cultivated cardoon, and wild cardoon C. cardunculus var. sylvestris, which cross fully. GP2 includes other Cynara species, such as C. syriaca, C. algarbiensis, C. baetica, and C. humilis. Crossing with them produces few seeds, and F₁ hybrids are usually sterile (Lanteri & Portis 2008).

Significance of the GP Concept for Breeding

The GP concept helps the breeder plan crosses and choose methods of gene transfer. If the desired trait is in GP1, crossing is not difficult. If it is in GP2, special methods are required, but transfer is possible. If it is in GP3, embryo rescue, bridge crosses, or genetic engineering are necessary.

In addition, the GP concept helps assess the risk of gene introgression from crop plants into wild populations. If a crop plant and a wild relative are in the same GP, gene flow is possible, which can lead to genetic contamination of wild populations.

Self-Incompatibility (SI)

Self-incompatibility is a mechanism that prevents self-pollination and promotes cross-pollination. Several types of SI are distinguished:

• Gametophytic self-incompatibility (GSI) — the reaction is determined by the pollen genotype. If the S-allele of the pollen matches one of the S-alleles of the pistil, the pollen tube does not grow. GSI is widespread in Rosaceae, Solanaceae, Plantaginaceae (Chabert & Mallinger 2025).
• Sporophytic self-incompatibility (SSI) — the reaction is determined by the genotype of the sporophyte (pollen grain). If the pollen carries at least one S-allele matching the pistil, a rejection reaction occurs. SSI is characteristic of Brassicaceae, Asteraceae, Convolvulaceae (Chabert & Mallinger 2025).
• Heteromorphic self-incompatibility (HSI) — associated with heterostyly (different lengths of style and stamens). It occurs in Primulaceae, Rubiaceae, Lythraceae (Chabert & Mallinger 2025).
• Late self-incompatibility (LSI) — the reaction occurs in the ovary, after pollen tube germination. It may be pre- or postzygotic. It occurs in many woody plants (Chabert & Mallinger 2025).
• Diallelic self-incompatibility (DSI) — described in Oleaceae, including olive. It is characterized by the presence of only two mutually incompatible phenotypes (Chabert & Mallinger 2025).

Early-Acting Inbreeding Depression (EID)

EID is inbreeding depression manifesting at early stages of embryo development. It leads to abortion of some embryos due to genetic load. EID is characteristic of self-compatible cross-pollinated species, such as blueberry Vaccinium spp. (Chabert & Mallinger 2025). In such cases, self-pollination leads to reduced seed and fruit set. To maximize yield, cross-pollination must be ensured.

Xenia

Xenia is the influence of the pollen donor on seed and fruit quality. Mechanisms of xenia include double fertilization, transposons, phytohormones, and small RNAs (Chabert & Mallinger 2025). Xenia is important for planning varietal mixtures: even in self-compatible and self-fertile crops, cross-pollination can improve product quality.

Evolutionary Bases of Reproductive Compatibility

Reproductive compatibility is an evolutionarily determined trait. It is related to phylogenetic proximity: the closer the species, the higher the probability of successful crossing. However, there are exceptions associated with polyploidy, chromosomal rearrangements, and genetic incompatibilities.

Understanding reproductive biology is necessary for planning crosses, selecting pollinator varieties, and assessing the risk of gene introgression.

Searching for Pathogen Resistance Donors

One of the main tasks of breeding is creating varieties resistant to diseases and pests. CWR are the most important source of resistance genes. For example, rust-resistance genes in wheat were transferred from Aegilops species. Late blight resistance genes in potato—from wild Solanum species. Nematode resistance genes in tomato—from S. peruvianum.

Phylogenetic analysis makes it possible to purposefully search for such genes. The sister groups of a crop are the first candidates. If a sister group has resistance, it can be transferred into the crop.

Expanding the Genetic Base of Crops

Modern varieties possess a narrow genetic base. This makes them vulnerable to new diseases and climatic changes. CWR help expand the genetic base by introducing new alleles. For example, for artichoke and cardoon, wild cardoon is a source of genetic diversity that can be used to improve the crops (Lanteri & Portis 2008).

Predicting Crossability and Compatibility

The GP concept and phylogenetic analysis make it possible to predict crossability. If a species is in GP1, crossing is not difficult. If in GP2, special methods are required. If in GP3, embryo rescue or genetic engineering is necessary.

In addition, SI, EID, and xenia must be taken into account. For crops with SI, compatible pollinator varieties must be selected. For crops with EID, cross-pollination must be ensured. For crops with xenia, pollen donors that improve product quality must be selected.

Understanding the Origin of Weedy and Feral Forms

Weedy and feral forms often originate from crop plants or their hybrids with wild relatives. Understanding phylogenetic relationships and hybridization processes helps predict the appearance of such forms and develop control measures. For example, red rice in cultivated rice fields is a form that arose as a result of hybridization and natural selection (Connor et al. 2011).

Breeding for Resistance to Climatic Stresses

Climate change makes the search for sources of drought tolerance and salt tolerance especially urgent. Xerophytes and salt-tolerant plants possess a number of adaptations: effective sequestration of Na⁺ in vacuoles, operation of NHX transporters and H⁺-pyrophosphatases, altered structure of ABA receptors, fewer stomata, but faster stomatal response (Chen et al. 2024). Phylogenetic analysis shows which of these traits are ancient and which arose independently in different lineages. This helps purposefully search for donors among CWR.

Use of Polyploidy in Breeding

Polyploidy is widely used to create new forms with valuable traits. Artificial polyploidization makes it possible to:

• restore fertility of sterile hybrids;
• increase organ size;
• increase resistance to stresses;
• create new forms with valuable traits.

For example, triploid forms of banana and watermelon possess increased productivity. Allopolyploids, such as Triticale (a hybrid of wheat and rye), combine valuable traits of both parents.

Phylogenetic Position and Systematics

Wheat belongs to the family Poaceae (grasses), order Poales, clade commelinids, Monocots (APG IV 2016; Judd et al. 2018). The genus Triticum includes several species, among which the most important cultivated ones are: T. aestivum (bread wheat), T. durum (durum wheat), T. dicoccum (emmer wheat), T. monococcum (einkorn wheat). Wild species: T. urartu, T. boeoticum, T. dicoccoides, and others.

Poaceae is one of the largest and economically most important families of angiosperms. Grasses dominate herbaceous communities, and their grain forms the basis of human nutrition. Wheat, rice, maize, and barley together provide more than half of world grain production (Connor et al. 2011). The family Poaceae is characterized by a number of specialized traits: hollow stems (culms), two-ranked leaves with a sheath and ligule, spikelet inflorescences, the caryopsis fruit, and wind pollination (Judd et al. 2018).

Polyploidy and Genome Structure

Wheat is a classic example of allopolyploidy. Triticum aestivum (bread wheat) is an allohexaploid with chromosome number 2n = 6x = 42 (Judd et al. 2018; Connor et al. 2011). Its genome consists of three subgenomes, designated A, B, and D, each of which originates from parental species:

• A genome — from Triticum urartu (wild einkorn species);
• B genome — from Aegilops speltoides (or a closely related species);
• D genome — from Aegilops tauschii (Tausch’s goatgrass).

Triticum durum (durum wheat) is an allotetraploid with chromosome number 2n = 4x = 28, containing the A and B genomes. Triticum dicoccum (emmer) is also an allotetraploid (AABB). Triticum monococcum is a diploid (2n = 2x = 14, A genome).

The scheme of origin of wheat allopolyploids includes several stages of hybridization and chromosome doubling. First, the diploid species T. urartu (AA) and Ae. speltoides (BB) hybridized, giving rise to the allotetraploid T. dicoccoides (AABB). Then this tetraploid crossed with Ae. tauschii (DD), which led to the emergence of the allohexaploid T. aestivum (AABBDD) (Judd et al. 2018). This process occurred under domestication and in nature.

Domestication and Evolution of Economically Valuable Traits

Wheat was domesticated in the Fertile Crescent (Near East) about 10,000 years ago. The first to be domesticated were diploid species (einkorn, T. monococcum) and tetraploid species (emmer, T. dicoccum). Hexaploid wheat T. aestivum arose later, probably already under agriculture (Connor et al. 2011).

Key traits of wheat domestication:

• Loss of grain shattering. In wild cereals, spikes disintegrate into individual spikelets at maturity, ensuring seed dispersal. In cultivated wheat, the spike remains intact until harvest. This trait is controlled by genes influencing formation of the abscission layer (Judd et al. 2018).
• Increase in grain size. During breeding, grain size increased significantly.
• Synchronous ripening. Cultivated forms ripen simultaneously, which facilitates harvesting.
• Loss of hulledness. In wild forms, the grain is tightly covered by glumes (flowering glumes); in cultivated forms, the grain is naked (in bread and durum wheat).
• Change in photoperiodic response. Cultivated forms are adapted to different latitudes.

Genetic progress in wheat yield in the 20th century was associated mainly with changes in harvest index—the proportion of biomass directed to grain. Breeding for short stature (semidwarf) made it possible to increase lodging resistance and redistribute assimilates in favor of grain. This became the basis of the “Green Revolution” (Connor et al. 2011). Comparison of historical varieties shows that total aboveground biomass changed little, but harvest index increased from 0.35 to 0.45–0.50. The photosynthetic capacity of leaves also did not increase—the main progress was achieved through redistribution of resources (Connor et al. 2011).

Modern wheat yields grow linearly, but relative growth rates are declining. From 1966 to 2006, the average increase was 39 kg/ha per year (Connor et al. 2011). Further increases in yield require new approaches, including the use of genetic resources of wild relatives.

Genetic Resources and Wild Relatives

Gene pools of wheat:

• GP1 (primary gene pool): T. aestivum, T. durum, T. dicoccum, T. monococcum, as well as wild forms T. urartu, T. boeoticum, T. dicoccoides. These species cross easily with cultivated wheat; hybrids are fertile.
• GP2 (secondary gene pool): other Aegilops species (for example, Ae. speltoides, Ae. tauschii, Ae. umbellulata). Crossing is possible but requires special methods.
• GP3 (tertiary gene pool): more distant cereal species (rye Secale, barley Hordeum, wheatgrass Elymus). Crossing requires embryo rescue, bridge crosses, or genetic engineering.

Triticale is an artificial allopolyploid obtained by crossing wheat (Triticum) and rye (Secale). It combines the high yield of wheat and the resistance of rye to unfavorable conditions (Connor et al. 2011). This is an example of the successful use of a distant relative in breeding.

Wild relatives of wheat are a source of genes for resistance to rust, powdery mildew, nematodes, as well as genes for drought tolerance and salt tolerance. For example, stem rust resistance genes (Sr genes) were transferred from Aegilops and Triticum species. Leaf rust resistance genes (Lr genes)—also from Aegilops. These genes are widely used in modern varieties.

Practical Significance for Breeding

For wheat breeding, it is important to:

• Use GP1 for rapid improvement of traits.
• Use GP2 for transferring disease-resistance genes.
• Use GP3 for transferring genes from distant relatives using biotechnology.
• Take into account the polyploid nature of wheat when planning crosses.
• Apply marker-assisted selection (MAS) for selection based on resistance and quality genes.
• Use genetic maps and QTL analysis to identify loci controlling yield and quality.

Phylogenetic Position and Systematics

Soybean Glycine max belongs to the family Fabaceae (Leguminosae), subfamily Faboideae (Papilionoideae), order Fabales, clade fabids, Rosids (APG IV 2016; Judd et al. 2018). Fabaceae is the third largest family of angiosperms, including about 750 genera and 19,500 species. Characteristic traits of the family: compound leaves with stipules and pulvini, the legume fruit, and the ability for symbiotic nitrogen fixation with nodule bacteria (Rhizobium) (Judd et al. 2018).

The subfamily Faboideae (Papilionoideae) is characterized by zygomorphic papilionaceous flowers (with a banner, wings, and keel), diadelphous stamens (9 + 1), and usually pinnate leaves. Soybean belongs to the tribe Phaseoleae, which also includes common bean (Phaseolus), cowpea (Vigna), hyacinth bean (Lablab), and other important crop plants (Judd et al. 2018).

The genus Glycine includes about 30 species distributed in Australia, East Asia, and Southeast Asia. Cultivated soybean G. max is an annual herbaceous plant. Wild relatives: G. soja (wild soybean), G. tabacina, G. tomentella, G. canescens, and others. G. soja is the closest wild relative of cultivated soybean, with which it crosses easily. G. max and G. soja are often considered one species, G. max (cultivated soybean) and G. max subsp. soja (wild soybean) (Judd et al. 2018).

Domestication and Evolution of Economically Valuable Traits

Soybean was domesticated in China about 3,000–5,000 years ago. Key traits of domestication:

• Loss of pod shattering. In wild soybean, pods crack at maturity and seeds disperse. In cultivated soybean, pods remain closed.
• Increase in seed size. Cultivated forms have larger seeds.
• Change in growth form. Wild soybean is a climbing plant; cultivated soybean is erect.
• Synchronous ripening.
• Loss of bitterness and antinutritional substances.
• Increased oil and protein content.

Soybean is predominantly self-pollinated (IB—inbreeder), which facilitates maintenance of varieties but limits genetic diversity. Nevertheless, there is significant diversity of landraces and wild forms (Connor et al. 2011).

Genetic Resources and Wild Relatives

Gene pools of soybean:

• GP1: G. max (cultivated soybean) and G. soja (wild soybean). They cross easily; hybrids are fertile.
• GP2: other Glycine species (for example, G. tabacina, G. tomentella). Crossing is possible, but with difficulties. Many of them are polyploids (2n = 40, 80 and more).
• GP3: more distant Fabaceae species.

Wild relatives of soybean are a source of genes for resistance to diseases (for example, soybean rust, powdery mildew, nematodes), as well as genes associated with oil and protein quality, drought tolerance, and salt tolerance. For example, G. soja possesses resistance to some pathogens that cultivated soybean lacks.

Polyploidy in Glycine

The genus Glycine includes diploid (2n = 20, 40) and polyploid (2n = 80) species. Cultivated soybean G. max is a diploid (2n = 40). The diploid species G. soja and G. max have 20 pairs of chromosomes. Polyploid species, such as G. tomentella (2n = 80), are allopolyploids and can be used in breeding to transfer resistance genes, although crossing with them is difficult (Judd et al. 2018).

Practical Significance for Breeding

For soybean breeding, it is important to:

• Use G. soja as a source of resistance and adaptability genes.
• Apply MAS for selection based on disease-resistance and quality genes.
• Take into account self-incompatibility and inbreeding depression when planning crosses.
• Use polyploid species for gene transfer using embryo rescue or genetic engineering.
• Expand the genetic base of varieties through wild relatives.

Phylogenetic Position and Systematics

The genus Brassica belongs to the family Brassicaceae (Cruciferae), order Brassicales, clade malvids, Rosids (APG IV 2016; Judd et al. 2018). Brassicaceae is one of the largest families of angiosperms, including about 338 genera and 3,710 species. Characteristic traits: herbs (less often shrubs) containing glucosinolates (mustard oils), flowers with 4 sepals, 4 petals (cruciform), 6 stamens (tetradynamous—2 short and 4 long), superior ovary with a false septum (replum), fruit a siliqua or silicula (Judd et al. 2018).

Brassicales is characterized by the presence of glucosinolates and myrosin cells. The order includes 19 families, among which Brassicaceae is the largest. The family Brassicaceae is monophyletic, which is confirmed by the presence of a false septum in the ovary and the siliqua, as well as by molecular data (Judd et al. 2018; APG IV 2016).

The genus Brassica includes several cultivated species:

• B. oleracea (garden cabbage): cabbage, kale, broccoli, cauliflower, Brussels sprouts, kohlrabi;
• B. rapa (turnip, Chinese cabbage): turnip, Chinese cabbage, bok choy;
• B. napus (rapeseed): rapeseed, canola;
• B. juncea (brown mustard): Chinese mustard, leaf mustard;
• B. nigra (black mustard): black mustard;
• B. carinata (Ethiopian mustard): Ethiopian mustard.

Wild relatives: B. rapa subsp. sylvestris, B. oleracea var. oleracea (wild cabbage), B. tournefortii, B. fruticulosa, B. spinescens, and others. Many of them are distributed in the Mediterranean and Western Asia (Judd et al. 2018).

Polyploidy and the “Triangle of U”

The genus Brassica is a classic example of allopolyploidy and hybridization. Three diploid species (B. rapa, B. oleracea, B. nigra) and three allotetraploid species (B. napus, B. juncea, B. carinata) form the so-called “Triangle of U,” described by the Japanese geneticist Nagaharu U in 1935. Although the term “Triangle of U” itself is not mentioned in the sources used, its content is confirmed by data on the genome structure of these species (Judd et al. 2018).

Genome formulas:

• B. rapa — AA (2n = 20);
• B. oleracea — CC (2n = 18);
• B. nigra — BB (2n = 16);
• B. napus — AACC (2n = 38) — hybrid B. rapa × B. oleracea;
• B. juncea — AABB (2n = 36) — hybrid B. rapa × B. nigra;
• B. carinata — BBCC (2n = 34) — hybrid B. nigra × B. oleracea.

These allotetraploids arose as a result of interspecific hybridization and chromosome doubling. They are fertile because each chromosome has a homolog for conjugation in meiosis. Allopolyploidy played a key role in the evolution and domestication of Brassica (Judd et al. 2018).

Domestication and Evolution of Economically Valuable Traits

Cabbage (B. oleracea) is one of the most striking examples of diversity arising as a result of selection for different organs:

• Kohlrabi — selection for stem tuber;
• Cabbage — selection for apical bud;
• Brussels sprouts — selection for lateral buds;
• Broccoli — selection for inflorescences;
• Cauliflower — selection for underdeveloped inflorescences;
• Kale — selection for leaves.

All these forms belong to one species B. oleracea, but were selected for different organs. This is an example of modularity of development: different parts of the plant can change independently (Ambrose & Ferrandiz 2013; Judd et al. 2018).

Brassica rapa includes turnip (root crop), Chinese cabbage (leaves), and bok choy (petioles). Brassica napus — rapeseed, the most important oilseed crop. Rapeseed oil is used for food and for biodiesel production. Canola is a rapeseed variety with low content of erucic acid and glucosinolates.

Key traits of Brassica domestication:

• Loss of bitterness. Glucosinolates impart a bitter taste. Selection for reduced glucosinolate content led to the creation of canola.
• Increase in organ size. Heads, root crops, inflorescences.
• Change in growth form. From wild herbaceous forms to cultivated ones.
• Synchronous ripening.
• Disease resistance.

Genetic Resources and Wild Relatives

Gene pools of Brassica:

• GP1: cultivated species and their wild forms of the same species (B. oleracea var. oleracea, B. rapa subsp. sylvestris). They cross easily.
• GP2: other Brassica species (for example, B. tournefortii, B. fruticulosa). Crossing is possible with difficulties.
• GP3: species of other Brassicaceae genera (for example, Sinapis, Raphanus, Eruca). Crossing requires special methods.

Wild relatives of Brassica are a source of genes for resistance to diseases (for example, clubroot Plasmodiophora brassicae, Alternaria blight, bacterioses), as well as genes associated with oil quality (erucic acid content, glucosinolates), drought tolerance, and salt tolerance. For example, clubroot resistance was transferred from B. rapa and B. oleracea into cultivated varieties.

Practical Significance for Breeding

For Brassica breeding, it is important to:

• Use allopolyploidy to create new forms with valuable traits.
• Apply MAS for selection based on oil quality and disease-resistance genes.
• Use wild relatives to expand the genetic base.
• Take self-incompatibility (SSI) into account when planning crosses. Sporophytic self-incompatibility is widespread in Brassica, preventing self-pollination (Chabert & Mallinger 2025).
• Use heterosis (F₁ hybrids) to increase yield.

Homology

Homology is the similarity of traits due to their origin from the same trait of a common ancestor. Homologous structures may perform different functions and differ externally, but their development is controlled by the same genes and regulatory networks (Judd et al. 2018).

A classic example in botany: all leaves of angiosperms are homologous as organs that arose from lateral primordia on the flanks of the meristem. Even if a leaf has turned into a tendril, spine, or trapping apparatus, it remains a homolog of a typical leaf. Similarly, all flowers of angiosperms are homologous as structures, although they may differ greatly in shape, size, and function.

Homology may be complete (organs as a whole are homologous) or partial (individual parts are homologous; for example, sepals and petals may be homologous within the ABC model, but differentiated). Homology may also be serial (for example, homologous segments in a flower—sepals, petals, stamens, carpels—arose as modified leaves).

Analogy

Analogy is the similarity of traits that arose independently in different lineages and was not inherited from a common ancestor. Analogous structures perform similar functions, but have different origins and often different genetic control (Judd et al. 2018).

A classic example in zoology: wings of birds and wings of insects. Both serve for flight, but arose independently and have completely different structure. In botany, analogous structures are, for example, cactus spines (modified leaves) and hawthorn spines (modified shoots). Both perform a protective function, but arose from different organs.

Parallelism

Parallelism is the independent evolution of similar traits in closely related species descending from a common ancestor that did not possess this trait. Parallelisms are often associated with the fact that closely related species have similar gene networks, and changes in the same genes can lead to similar phenotypes (Judd et al. 2018).

Example: fleshy fruits in different genera of Solanaceae (tomato, pepper, eggplant). All of them originated from a common ancestor with dry fruits, but independently acquired fleshy fruits. This is parallelism, not homology, because the trait arose independently in each lineage.

Convergence

Convergence is the independent evolution of similar traits in distant, unrelated species. Unlike parallelism, convergence occurs in species that are not close relatives and often have different genetic control of the trait.

Example: succulence as an adaptation to arid conditions arose independently in cacti (Cactaceae), spurges (Euphorbiaceae), stonecrops (Crassulaceae), and other families. All of them accumulate water in tissues, but the mechanisms and genetic control may differ (Chen et al. 2024).

Phylogenetic Context

The first and main criterion is phylogenetic. If a trait is present in sister groups and in their common ancestor, it is probably homologous. If a trait is present only in one of the sister groups or arose independently in several unrelated lineages, it is probably analogous (Judd et al. 2018; Pandey 2021).

To reconstruct the evolution of a trait, it is necessary to map it onto a phylogenetic tree and determine at which nodes changes occurred. Methods of ancestral state reconstruction (parsimony, maximum likelihood, Bayesian) make it possible to estimate what state the ancestor had and where transitions occurred (Judd et al. 2018).

Developmental Data (Evo-Devo)

The second criterion is developmental data. If two traits are controlled by the same genes and regulatory networks, they are probably homologous. If the control differs, the trait may be analogous (Ambrose & Ferrandiz 2013).

Modern evo-devo has shown that many morphological innovations arise not through the appearance of new genes, but through changes in the expression of existing genes—their co-option into new contexts. For example, MADS-box genes controlling flower development were co-opted for new functions in different lineages. This means that similar traits can be controlled by the same genes, but in different combinations, which complicates the distinction between homology and analogy (Ambrose & Ferrandiz 2013).

Deep Homology

Deep homology is a phenomenon in which similar structures are controlled by the same ancient genes, even if the structures themselves are not homologous. For example, eyes in different animals are controlled by the gene Pax6, although the eyes of insects and vertebrates arose independently (Judd et al. 2018).

In plants, an example of deep homology may be the control of leaf and flower development by the same MADS-box genes. Although leaves and flowers are not homologous as organs, their development is controlled by overlapping gene networks (Ambrose & Ferrandiz 2013).

Modularity and Co-Option

Modularity is a property of developmental systems that allows different parts of the organism to change independently. Modularity reduces constraints and allows the developmental system to explore new morphospace (Ambrose & Ferrandiz 2013).

Co-option is the recruitment of an existing gene or pathway for a new function. For example, the LEAFY gene in pea was co-opted for the development of compound leaves, although initially it controlled flower development. Co-option can lead to the appearance of similar traits in unrelated species, which complicates the distinction between homology and analogy (Ambrose & Ferrandiz 2013).

Fleshy Fruits

Fleshy fruits are one of the most striking examples of parallel evolution. They arose independently in different families of angiosperms:

• Solanaceae: tomato, pepper, eggplant, physalis;
• Rosaceae: apple, pear, plum, cherry, raspberry, blackberry;
• Vitaceae: grape;
• Moraceae: fig, mulberry;
• Myrtaceae: guava, feijoa, clove tree;
• Cactaceae: prickly pear, dragon fruit.

All these fruits perform a similar function—attracting animal dispersers. However, they arose independently and have different anatomical structure. For example, the fleshy part of the tomato fruit is formed from the mesocarp and endocarp, while in apple it is formed from the hypanthium (enlarged receptacle). This is an example of convergence, not homology (Judd et al. 2018).

From the standpoint of breeding, this means that traits of fleshiness in different crops are controlled by different genes, and their transfer between distant species may be difficult. However, within one family (for example, Solanaceae), fleshiness may be controlled by similar genes, which facilitates breeding.

Large Seeds

Large seeds are another example of parallel evolution. An increase in seed size occurred independently in many crops:

• Cereals: wheat, rice, maize, barley;
• Legumes: soybean, common bean, pea, lentil;
• Oilseeds: sunflower, rapeseed, peanut;
• Vegetables: pumpkin, cucumber, watermelon.

The increase in seed size is associated with changes in regulatory networks controlling cell division and expansion, as well as with changes in the metabolism of storage substances. Although the trait is phenotypically similar, genetic control may differ in different crops (Ambrose & Ferrandiz 2013; Connor et al. 2011).

For example, in wheat, the increase in grain size is associated with changes in genes controlling cell division in the endosperm. In soybean—with changes in genes controlling oil and protein synthesis. These are examples of parallelisms within one family and convergence between different families.

Loss of Shattering

Loss of fruit and seed shattering is a key domestication trait in many crops. In wild cereals, spikes disintegrate into individual spikelets at maturity, ensuring seed dispersal. In cultivated forms, the spike remains intact until harvest.

Loss of shattering occurred independently in different crops:

• Wheat: mutations in genes controlling formation of the abscission layer (Q locus, Br genes);
• Rice: mutations in the genes sh4 and qSH1;
• Maize: mutations in genes controlling ear formation;
• Sorghum: mutations in genes controlling panicle formation.

Although the trait is phenotypically similar, genetic control differs. This is an example of parallel evolution at the phenotypic level, but convergent at the genetic level. A similar phenotype is achieved by different paths (Judd et al. 2018; Connor et al. 2011).

For breeding, this means that when transferring the trait of loss of shattering from one species to another, it is necessary to take into account that the genes may be different, and simple crossing may not give the expected result.

Loss of Bitterness

Loss of bitterness occurred independently in many vegetable crops:

• Cucumber: loss of cucurbitacins;
• Cabbage: reduction of glucosinolate content (creation of canola);
• Eggplant: reduction of steroidal glycoalkaloid content;
• Pumpkin: reduction of cucurbitacin content.

In cucumber, loss of bitterness is associated with mutations in genes of cucurbitacin biosynthesis (Bi, Bt genes). In cabbage—with mutations in glucosinolate biosynthesis genes. This is an example of convergence at the phenotypic level and parallelism within one family (Judd et al. 2018).

Succulence

Succulence is an adaptation to arid conditions that arose independently in cacti (Cactaceae), spurges (Euphorbiaceae), stonecrops (Crassulaceae), ice plants (Aizoaceae), and other families. All these plants accumulate water in tissues, but the mechanisms and genetic control differ (Chen et al. 2024).

Studies show that in xerophytes and salt-tolerant plants, the number of copies of NHX and H⁺-PPase genes is often increased, the structure of ABA receptors PYR/PYL is altered, stomatal density is reduced, but the speed of their response is increased. These adaptations arose independently in different lineages, which indicates convergent evolution at the molecular level (Chen et al. 2024).

Tuber Formation

Tuber formation is a trait that arose independently in potato (Solanum tuberosum), sweet potato (Ipomoea batatas), yam (Dioscorea spp.), Jerusalem artichoke (Helianthus tuberosus), and other crops. In potato, tubers are modified underground shoots (stolons); in sweet potato, they are modified roots. This is an example of analogy: different organs perform a similar function (Connor et al. 2011).

Artichoke and Cardoon: Parallel Domestication

Artichoke (Cynara cardunculus var. scolymus) and cardoon (C. cardunculus var. altilis) originate from wild cardoon (C. cardunculus var. sylvestris). Domestication proceeded independently: artichoke was selected for large flower heads, cardoon for fleshy petioles. Genetic studies have shown that these two forms are significantly differentiated, while a high level of diversity is preserved within cultivated forms (Lanteri & Portis 2008).

This is an example of parallel domestication within a single species complex. The traits selected in artichoke (large flower heads) and cardoon (fleshy petioles) are controlled by different sets of genes, although both traits are associated with an increase in organ size.

Genes Controlling Morphological Innovations

Modern studies show that many morphological innovations in crop plants are associated with duplications and diversification of MADS-box genes. For example, duplications in the AP1/FUL and AP3/PI lineages correlate with the emergence of new flower morphologies in core eudicots (Ambrose & Ferrandiz 2013).

In wheat, duplications of the AP3 and PI genes are associated with differentiation of the perianth. In maize, duplications of AP3 genes are associated with differentiation of lodicules and lemmas. These examples show that parallel morphological changes may be associated with changes in the same gene families, but in different combinations (Ambrose & Ferrandiz 2013).

Strategy for Searching Donors

If a trait is homologous, it can be sought in close relatives. If it is analogous, the search should be directed toward species with an independent origin of the trait, which may be more difficult (Judd et al. 2018).

For example, rust resistance in wheat is homologous in different Triticum and Aegilops species, so it can be transferred through ordinary crosses. Drought resistance in xerophytes arose independently in different lineages, and its transfer may require different genes in each case (Chen et al. 2024).

Predicting Crossability

Homologous traits are often associated with close kinship, which increases the probability of successful crossing. Analogous traits may be present in distant species, crossing with which is difficult (Judd et al. 2018; Connor et al. 2011).

Choosing Gene Transfer Methods

If a trait is homologous, traditional breeding methods can be used. If it is analogous, MAS, embryo rescue, bridge crosses, or genetic engineering may be required (Connor et al. 2011).

Understanding the Limits of Breeding

Some traits may be incompatible with each other due to pleiotropic effects or developmental constraints. Understanding the evolutionary history of traits helps predict such constraints. For example, breeding for large fruit in tomato led to the loss of salt tolerance, which is associated with pleiotropic effects of genes controlling fruit size (Chen et al. 2024).

Using Genomic Data

Modern genomic studies make it possible to identify genes controlling economically valuable traits and compare them in different species. This makes it possible to distinguish homology and analogy at the molecular level and purposefully use genetic resources (Ambrose & Ferrandiz 2013; Chen et al. 2024).

Phylogenetic Reconstruction

The first step is constructing a phylogenetic tree based on molecular and morphological data. Then trait states are mapped onto the tree and their evolution is reconstructed (Judd et al. 2018; Pandey 2021).

Comparative Transcriptomics

Comparison of gene expression in different species makes it possible to identify genes involved in the development of similar traits. If the genes are the same—this indicates homology or deep homology. If different—analogy or convergence (Ambrose & Ferrandiz 2013).

Functional Genetics

Experimental verification of gene function using mutagenesis, transgenesis, or genome editing (CRISPR/Cas9) makes it possible to confirm the role of specific genes in trait development. This is necessary for distinguishing homology and analogy at the functional level (Ambrose & Ferrandiz 2013).

Paleobotany

Fossil remains make it possible to trace the evolution of traits over time and determine when and where they arose. This is especially important for distinguishing homology and analogy in extinct groups (Judd et al. 2018).