Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Mexico as a nation is not conceivable without the existence of maize. For most Mexicans, more than being domesticated, maize was handcrafted through thousands of years of ancestral work. The plant has been at the heart of cultural creativity for all generations and remains an essential element of popular Mexico. As described by Guillermo Bonfil (1982), the establishment of maize as the fundamental Mexican crop required the design and implementation of the technical procedures necessary for its cultivation and consumption, including the specific organization of time and space in response to a physiological rhythm imposed by the plant. It also guided the emergence of immensely diverse religious practices, allowed the development of a unique culinary art, and became the reference necessary to understand the popular organization of the Mexican life style. In contrast to developed countries where it is fundamentally used for agro-industrial or animal production, maize in Mexico is mainly cultivated for human nutrition, in all states, and under a wide range of climatic conditions. Its consumption represents the main source of protein and energy in rural regions, particularly in the poorest communities. Corn is at the table of millions of Mexicans daily, either in the form of a wide diversity of regional meals, or simply as its most important derivative, the tortilla.

Maize was domesticated from its wild progenitor teosinte (derived from “ teoc-intli” in nahuatl language: “ teotl” = sacred and “ cintli” = dried ear of corn), a common name given to a group of annual and perennial species of the genus Zea native to Mexico and Central America (see Chap. 20; reviewed in Doebley, 2004; Matsuoka, 2005). An extensive phylogenetic analysis based on molecular comparisons of landraces and wild taxa indicates that maize arose from a single domestication event occurred in southern Mexico about 9,000 years ago (Matzuoka et al., 2002). Domestication resulted in a group of ancestral landrace varieties that subsequently spread throughout the continent and eventually to the rest of the World, adapting to human practices in different ecological and geographical regions where the crop has been established. As the main center of origin and domestication, Mexico has the largest diversity of maize genetic resources. Through continuous divergent selection, the plant has diversified into a multitude of populations with distinct adaptations to climate, latitude, and length of growing season, and has specialized to fulfill a wide range of specific human purposes (Hemández, 1985; Ortega Paczka, 2003).

In this chapter we review our current knowledge of the genetic variability found in Mexican landraces, and discuss their most recent classification on the basis of morphology, geographical distribution, and some biochemical characteristics. We also anticipate some of the future implications that landrace genome characterization will have for the understanding of the molecular and functional diversity of maize.

2 Conservation of Mexican Maize Germplasm

There is a substantial confusion in the terminology used to refer to maize germ-plasm. As indicated by Ortega Paczka (2003), the term “ variedad criolla” or “ criollo” is commonly used by many Mexican agronomists or breeders to name native local populations for which a pedigree is absent. The term “ criollo” was used to name people from Spanish descent that were born in the Americas, and therefore has little in common to maize native populations maintained by Mexican farmers.

The term “ raza” (race or landrace) should be in principle less confusing, as it is generally accepted that a landrace is a population of individuals that share a large number of genetically inherited traits that allow a clear and consistent phenotypic distinction from other maize populations. The name of a landrace can sometimes refer to an obvious phenotypical trait such as “ Cónico” for a conical ear shape, or “ Negrito” for prevalent seed color. In other cases it refers to a location where the germplasm was initially collected or is predominant ( Tuxpeño or Mixteco), or it can also reflect indigenous names given by local populations in their native language ( Nal Tel or Dzit Bacal).

In practice, the identification and description of a maize landrace is not obvious. In contrast to modern breeding schemes used by seed companies, traditional maize improvement is mostly empirical and relies on complex factors that include elements of tradition, intuition, affection, and improvisation. Since cross-pollination prevails as a reproductive habit, maize heterozygosity is usually high and advantageous in common selection procedures followed by rural communities. Owing to these practices, what is generally observed in Mexican local populations is a continuous variation in quantitative traits such as plant height, ear size, kernel row number, or flowering time. Additional traits such as seed color are also highly variable among individuals of a same landrace. The majority of native maize planted to date is the result of hybridization between several landraces, and often between native germplasm and modern varieties recently introduced through neighboring practices or human migration (Bellón et al., 2003; Bellón and Berthaud, 2004; Pressoir and Berthaud, 2004a, b).

While the effort of creating and preserving Mexican maize diversity is intimately related to social and economic activities of indigenous communities, peasants, and farmers, it is the enthusiasm and dedication of a relatively small group of scientists that has enable the systematic identification, description, and conservation of present collections of maize germplasm. Although a general interest for maize diversity dates back to pre-Colombian times, the systematic study and collection of maize germplasm initiated in the twentieth century, with the work of Chávez (1913), Vavilov (1931), Anderson and Cutler (1942), and Anderson (1946). The work of Kuslehov (1930) was essential to demonstrate for the first time that the variability of seed morphology is exceptionally high in Mexico. Based on these pioneering initiatives, from 1938 to 1951 the Mexican Ministry of Agriculture supported important efforts to identify the best native materials that could serve as parental lines for a national program of maize improvement. Although the majority of the results were never published, these efforts indirectly derived in a successful partnership between the Ministry of Agriculture and the Rockefeller Foundation, giving rise to the most valuable ex situ collections of Mexican maize germplasm existing to date: the collection kept at the Centro Internacional de Mejoramiento de Maiz y Trigo (CIMMYT), and the Germplasm Bank of the Instituto Nacional de Investigaciones Forestales y Agropecuarias (INIFAP). With more than 11,000 samples, the latter represents the most important collection of Mexican maize germ-plasm; a substantial portion has been duplicated and transferred both to CIMMYT and the National Seed Storage Laboratory (USDA-NSSL) in Forth Collins, Colorado. Unfortunately, the INIFAP infrastructure is from 1950 and a limited budget impedes renovation. Additional collections include the Native Species Germplasm Bank at the Autonomous University of Chapingo (Texcoco, México) with 2,500 samples from native local populations, and the collection of Colegio de Postgraduados (Montecillos, México) with 4,000 samples. CIMMYT international accessions total more than 22,000 and according to official authorities new introductions are constantly being added from cooperative regeneration projects around the world. Passport data on CIMMYT maize germplasm bank accessions have been compiled and is available on CD-ROM (for details see www.cimmyt.cgiar.org). To safeguard CIMMYT collections, duplicate samples of about four fifths are kept at the USDA-NSSL.

3 Identification and Classification of Mexican Landraces

Maize biologists do not always agree in the total number of maize landraces that exist in Mexico (Wellhausen et al., 1951; Sánchez et al., 2000; Ortega Paczka, 2003; Ron-Parra et al., 2006). The classical monograph published by Wellhausen et al. (1951) has been an essential reference for all subsequent reports. Based on general architecture, kernel cytological traits, and physiological characteristics (time of flowering, yield, and disease resistance), they classified 25 landraces into 5 major groups. The first one included four Ancient Indigenous Races believed to have arisen from a primitive pod corn ( Palomero Toluqueño, Arrocillo Amarillo, Chapalote, and Nal Tel). Other groups included landraces believed to be introduced from other regions of Central or South America (Pre-Colombian Exotic Races), or to have arisen after hybridization of ancient native and introduced landraces with teosinte (Prehistoric Mestizos); the fourth and fifth groups were considered to be composed of Modern Incipient Races or simply Poorly Defined Races.

A renewed interest for developing comprehensive systems of classification of Mexican landraces was initiated by Goodman (1972), Goodman and Bird (1977), and Cervantes et al. (1978) with the use of numerical techniques to classify morphological traits and geographical distribution. These studies resulted in a first set of well-defined racial groups that generally followed the relationships proposed by Wellhausen et al. (1951). Subsequent studies emphasized maize cytogenetic characteristics to report an extensive analysis of the chromosome constitution and geographic distribution of knob complexes (Kato, 1976, 1984; McClintock et al., 1981). These authors showed that specific knobs in chromosome 1, a small knob in the short arm of chromosome 3, and a large knob in the short arm of chromosome 7 are only distributed in landraces of the Pacific Coast, showing for the fist time a genetic link between primitive landraces such as Chapalote, Nal Tel, Harinoso del Ocho, Reventador and Zapalote Chico, and their relatives Blandito de Sonora, Cristalino de Chihuahua, Dulcillo de Sonora, Palomero Toluqueño, and Onaveño.

One of the first systematic attempts to biochemically characterize the genetic diversity of Mexican landraces was conducted by Doebley et al. (1985), who analyzed isozyme variation corresponding to 23 loci in 34 landraces. Their results indicated that the levels of variation in maize landraces are as high as in the teos-intes, demonstrating that maize is among the most genetically variable crops. Although they could not identify well-defined racial complexes, their study allowed the recognition of three groups of morphologically similar landraces: (1) The high elevation Mexican Pyramidal group, represented by Palomero Toluqueño, Cónico, Chalqueño, and Cacahuacintle, (2) The Northern and Northwestern group, represented by Apachito, Azul, and Gordo, and (3) The remaining races, including the Southern and Western low-elevation dent and floury landraces.

Additional identification of Mexican landraces has been reported by Ortega Paczka (1979), Benz (1986), and Sánchez (1989). After establishing the existence of two racial complexes (the “Mexican Pyramidal Ear Complex” and the “Mexican Narrow Ear Complex”), Benz (1986) proposed five more types ( Chatino Maizón, Choapaneco, Mixeño, Mixteco, and Serrano Mixe); however, a published illustration and description of these landraces has not been available, and their distinction from previously established accessions has been sometimes questioned (Ortega Paczka, 2003). Using morphological traits in combination to numerical taxonomy, Sánchez (1989) was able to describe four of the “poorly defined races” mentioned by Wellhausen and co-workers ( Blandito de Sonora, Dulcillo del Noroeste, Mushito, and Zamorano Amarillo), and confirmed the existence of three types identified by Ortega Paczka (1979): Coscomatepec, Motozinteco, and Elotero de Sinaloa.

Following these previous reports, and to assess the overall genetic diversity of Mexican germplasm, Sánchez et al. (2000) collected accessions from 50 landraces and analyzed their isozymatic and morphological characteristics under diverse growing conditions and locations. Although a very high level of variability was found among and within landraces, more than 65% of the alleles were present at frequencies below 1%, with some populations having low levels of genetic diversity, particularly those corresponding to rare variants planted in small fields in which seed for the next cycle comes from a small number of ears. Although Ortega Paczka et al. (1991) recognized the existence of at least 41 landraces, the study of Sánchez et al. (2000) reports the largest number of classified landraces on the basis of both morphological and biochemical evidence.

4 The Classification of Sánchez et al. (2000)

The most complete classification was proposed by Sánchez et al. (2000) and includes 50 out of 59 landraces divided in 5 major subdivisions (Fig. 1 and Table 1) on the basis of biochemical and morphological markers (Ron-Parra et al., 2006):

  • The Central and Northern Highlands Group is composed of 15 landraces. With one exception ( Dulce de Jalisco), all landraces of this group grow at elevations higher than 2,000 m. All group members are characterized by a low frequency of tassel branches, a weakly developed root system, and strongly pubescent leaf sheaths often pigmented by anthocyanins. Apachito, Azul, Gordo and Cristalino de Chihuahua are restricted to Northwest highlands at 2,000–2,600 m. These landraces have short plants (140–190 cm), early flowering (53–68 days), and rounded kernels. The group also includes the large Cónico subgroup ( Arrocillo Amarillo, Cacahuacintle, Chalqueño, Cónico, Cónico Norteño, Elotes Cónicos, Dulce de Jalisco, Mixteco, Mountain Yellow, Mushito, Negrito, Palomero de Chihuahua, Palomero de Jalisco, Palomero Toluqueño) showing conically shaped ears and high kernel row number (14–18).

  • The Eight-Rowed Group includes 12 landraces distributed at elevations comprised between 1,000 m and 1,800 m, with some along the Pacific Coast growing at elevations comprised between 100 m and 500 m ( Ancho, Blandito de Sonora, Bofo, Bolita, Elotes Occidentales, Harinero de Ocho, Jala, Onaveño, Tablilla de Ocho, Tabloncillo, Tabloncillo Perla, and Zamorano Amarillo). These landraces are 200–250 cm in height, and characterized by early-to-medium maturity cycles in which flowering is reached 70–80 days following germination.

  • The Late Maturity Group includes seven landraces. Whereas Comiteco, Coscomatepec, Motozinteco, Olotón and Tehua are all from Southern Mexico and distributed at elevations comprised between 900 m and 2,200 m, Dzit Bacal and Olotillo are from Southern States but cultivated at elevations between 500 m and 700 m. All are characterized by tall plants (290–380 cm), late-maturity cycles (88–110 days to flowering); many tassel branches, and high sensitivity to photoperiod and temperature fluctuations. Ears from Olotillo and Dzit Bacal have 8 rows; all others have 10–16 rows.

  • The Tropical Dents are 12 landraces subdivided in two subgroups. The first one includes Conejo, Nal Tel, Ratón, Tepecintle, Zapalote Chico and Zapalote Grande. All are characterized by short fast maturing plants adapted to low elevations. The second subgroup consists of Celaya, Nal Tel de Altura, Pepitilla, Tuxpeño, Tuxpeño Norteño, and Vandeño; all races are adapted to low to medium elevations (0–1,700 m) and characterized by tassel branched plants with long cylindrical ears (12–16 rows) and deeply dented kernels.

  • The Chapalote Group includes only four landraces ( Chapalote, Dulcillo del Noroeste, Elotero de Sinaloa, and Reventador). Although the isozymatic data had a tendency to suggest similarities with Tropical Dents and the Late Maturity groups (Fig. 1), members of the Chapalote group were classified in a separate class mainly because of their distinctive morphological characteristics. Chapalote is one of the most distinctive and ancient landraces of Mexico (Wellhausen et al., 1951). These races have slender cigar-shaped ears, weak tapering at both ends, and are commonly found at low elevations (100–500 m in the Pacific Coast).

Fig. 1
figure 1_27

Dendogram of 50 Mexican landraces based on morphological variables and isozyme frequencies (adapted from Sánchez et al., 2000)

Full size image
Fig. 2
figure 2_27figure 2_27figure 2_27figure 2_27

Examples of ears from 57 Mexican landraces. Groups correspond to the classification of Sánchez et al. (2001). A detailed description of these landraces and their distribution is presented in Table 1

Full size image
Table 1 General characteristics of 57 Mexican landraces
Full size table

Although no morphological data was available for some of the additional lan-draces, passport data was used to place Palomero de Jalisco, Mountain Yellow, Mixteco and Negrito in the Highlands Group, with the Cónico complex. With the same procedure, Mixeño and Negro de Chimaltenango were assigned to the Late Maturity Group, and Choapaneco to the Tropical Dents (Sánchez et al., 2000).

5 Genetic Erosion of Mexican Maize Diversity

Genetic erosion refers to germplasm and gene loss by the elimination of native local populations caused by exogenous factors such as the adoption of hybrid commercial varieties or drastic changes in land use (Plucknett et al., 1992). Several factors have eroded the endemic diversity of Mexican landraces. Over the last 70 years, the adoption of hybrid varieties has been particularly strong in regions like Central Mexico, the States of Jalisco and Nayarit, or Northeastern regions such as the State of Tamaulipas and the Laguna basin in Coahuila (Ron-Parra et al., 2006). The global economic integration of Mexico, in particular through the North American Free Trade Agreement (NAFTA), has increased pressure on agro-biodiversity, promoting monoculture specialization on farmers that now focus on high-yield varieties. At the same time, the Mexican government has constantly discouraged the production of colored maize by subsidizing the production of white corn, a policy that almost caused the extinction of pure landraces such as Tehua or Jala. According to Ortega Paczka (2003), there is at least one nearly extinguished landrace ( Tehua), 6 that are endangered ( Jala, Zamorano Amarillo, Vandeño, Zapalote Grande, Pepitilla, and Motozinteco), and 13 that are rare or difficult to find under cultivated conditions ( Apachito, Palomero Toluqueño, Palomero de Chihuahua, Mountain Yellow, Gordo, Blandito de Sonora, Tabloncillo, Choapaneco, Tepecintle, Nal Tel, Mixeño, Serrano Mixe, and Coscomatepec).

6 Landrace Genome Sequencing and Functional Maize Diversity

The phenotypic and molecular diversity of maize has been essential to harness important traits for crop improvement. On the basis of landrace germplasm, the activity of modern plant breeders gave rise to inbred lines currently used in hybrid production, causing significant improvements in yield, grain quality, resistance to biotic or abiotic stress, and maturity (Walden, 1979). For example, it is well known that members of the Tropical Dents such as Tuxpeño and Tuxpeño Norteño were fundamental for several breeding programs in Mexico and around the world. Interestingly, and despite the importance of selection-dependent bottleneck effects that drastically reduced genetic diversity, most maize genes have retained high levels of nucleotide diversity as compared to other cereals (Tenaillon et al., 2001; Wright et al., 2005; Yamasaki et al., 2005). Wright et al (2005) compared SNP diversity between maize inbred lines and teosintes and concluded that the number of genes that show signs of human selection was close to 1,200. This estimation is in agreement with subsequent studies indicating that less than 3% of maize genes have been the targets of intentional or unintentional modification of individuals in a population through human action (artificial selection; Yamasaki et al., 2007). All other genes remain unselected but show evidence of a population bottleneck associated with domestication and crop improvement. It is currently estimated that selected alleles at loci regulating plant architecture and seed nutritional quality were genetically fixed at least 4,400 years ago (Jaenicke-Depres et al., 2003).

A genome wide survey of gene content in B73 and Mo17 revealed that more than 20% of gene fragments examined in allelic contigs were not shared between these two inbred lines (Morgante et al., 2005), a divergence that can be largely due to the activity of transposable elements such as helitrons (Fu and Dooner, 2002; Lai et al., 2005). Considering that single nucleotide polymorphisms that distinguish two maize inbred lines are on average as significant as those distinguishing humans form chimpanzees (Tenaillon et al., 2001), reasonable predictions anticipate that the genomic divergence between two landraces is far more important. The activity of specific families of transposable elements diverge in maize landraces. For example, transcriptionally active MuDR elements, the regulatory element of the Mutator transposon family that had been found only in the two specific maize lines, is also present in specific accessions of Zapalote chico (Gutiérrez-Nava et al., 1998). It is likely that the large non-homologies that characterize maize genomes will substantially contribute to landrace structural diversity (Fu and Dooner, 2002). As a consequence, large-scale sequencing efforts concentrated in B73 will not be sufficient to fully understand maize genome organization and identify all genes. To complement the large-scale B73 sequencing initiative and explore landrace genomic diversity, we recently undertook the structural and functional characterization of the Palomero Toluqueño genome after estimating its size at ~,950 Mb (Mexican Maize Genome Team, unpublished results). A total of 1.2 million Sanger reads (10% HCot; 90% enzyme-based methyl-filtration) and 213 pyrosequencing runs (50% methyl-filtered, 50% whole genome sequencing) were sequenced at the National Laboratory of Genomics for Biodiversity (Langebio). The total sequence generated represents coverage of more than 3X the full genome; it has been complemented by in-depth pyrosequence-based global transcriptional analysis of the same landrace. As expected, at least 15% of codifying transcripts are not reported in publically available databases (Julio Vega-Arenguín et al., unpublished results), suggesting that a large portion of the molecular and functional diversity contained in Mexican landraces remains unexplored; the process of assembly and annotation is well underway.

Access to large-scale structural genomic information is rapidly transforming QTL mapping, positional cloning, and association approaches by allowing dissection of complex traits down to the gene or nucleotide level (Buckler et al., 2006). Since the resolution of association mapping is dependent upon the structure of linkage disequilibrium (LD or the non-random association of alleles between loci; Yu and Buckler, 2006) and does not require the generation of a mapping population, the diversity of Mexican landraces might represent the most suitable maize germ-plasm for these approaches. Whereas LD can be over 100 Kb for commercial elite inbred lines, it extends less than 1,000 bp for maize landraces (Buckler and Gore, 2007; Remington et al., 2001). Although the local prevalence of genomic non-ho-mologies might result in potential difficulties by reducing recombination and preserving LD, the access to large-scale landrace sequence information will represent an invaluable source of polymorphic information for exploring maize natural variation and exploiting allele diversity and recombination. We expect that a renewed interest in landrace germplasm will emerge with the development of new international initiatives to explore the functional diversity of maize.