Introduction

Mexico is the center of domestication and a center of diversity for maize (^{Matsuoka et al. 2002}; ^{Doebley 2004}), where it was domesticated from teosinte almost 9 000 years ago (^{Doebley, 2004}; ^{Kato et al, 2009}). Studies on maize variability show that there is great diversity in Mexican populations, besides the existence of teosinte, a wild relative from which cultivated maize descends (^{Sánchez et al, 1998}). In the second half of the last century great efforts were made to collect, conserve and use plant genetic resources and as a result, the genetic variability of the cultivated forms of maize is well represented in germplasm banks (^{Wellhausen et al, 1952}; ^{Hernández, 1985}).

Currently, there are about 60 races of maize in Mexico (^{Hernández, 2000}). The domestication of maize reduced the diversity of the species, and consequently the variation in domesticated maize populations has been reduced or restructured by genetic drift and selection (^{Vigouroux et al., 2002}). As a result of domestication, there is currently a large number of native populations adapted to specific environmental conditions and suited to a wide variety of uses (^{Warburton et al., 2008}). A wide number of landraces are used by indigenous farmers in Mexico, including a wide variety of native materials, regularly called landraces, or pure races of native maize or in hybridization with improved varieties, related to beans, squash and other crops (^{Dempsey, 1996}; ^{Ribeiro, 2004}).

In general, the loss of genetic diversity is caused by an increasing integration of producers into the market, imports and migration (^{Dyer and Qualset 2000}, ^{Van Dusen and Taylor 2005}), competition between modern and traditional varieties (^{Perales et al., 2005}), the type of land tenure (^{Dyer and Qualset, 2000}) and the progress of the human population which has led to environmental degradation (^{Ortega et al., 1999}).

The state of Nuevo León is known for its diversity of maize germplasm (^{Wellhausen et al, 1952}), where several accessions of landraces have been collected, such as Tuxpeño, Cónico Norteño, Tabloncillo, Tablilla de Ocho (Cárdenas 1995, cited by ^{Turrent and Serratos, 2004}) and Celaya (^{Ortega, 2007}). However, the number of races is low compared to that reported for other regions of Mexico (^{Hernández, 2000}). In general, the native varieties of maize in Mexico have been threatened by many factors, including the introduction of improved materials into the valleys of intensive production, the incidence of epiphytes, exposure to climatic factors, substitution by forage crops, in addition to demographic factors as migration of people, as a result only elderly people are familiar with this plant genetic resource, who in the medium term will have no one to whom transfer the germplasm and knowledge of how to grow it (^{Ortega, 2007}).

According to ^{Hoyt (1992)}, world population growth is projected to increase and also important climatic changes may disrupt habitats and modify agriculture. Therefore, efforts are needed to conserve and manage genetic resources and make them available for use. Based on the above, this research was conducted, in order to collect seed samples of native maize germplasm in the state of Nuevo León, for ex situ conservation and future use in different research programs.

The collection was carried out in the rainfed production system in 32 municipalities of the state of Nuevo León with significant records in diversity of native maize varieties, corresponding to regions of the Great Plains of North America, Northern Gulf Coastal Plain and Sierra Madre Oriental. Nuevo León is geographically located between 27° 49' and 23° 11' north latitude and 98° 26' and 101° 14' west longitude. The predominant climates are very warm dry steppe and extreme Bs (w1) (i') w'', with little rainfall throughout the year, semi-warm, extreme and semi-arid Bs (w1) (i') w'', with little rainfall throughout the year and cold sub-humid temperate Cb (w1) (i') w'', with summer rains (^{Medina García et al., 1998}). Soils are constituted by several associations of Calcaric Regosol, Calcaric Phaeozem, Rendzina with Chromic Luvisol and also Calcic Kastanozem soils in lytic phase (^{SPP, 1981}).

The collection was conducted during February and March 2008, after crop harvest. In each of the 32 municipalities maize samples were obtained directly from the farmer plot or storage place. From 20 to 50 ears per sample were collected and in some instances less ears were collected due to lack of availability, some collection was performed in grain, especially when the producer had only shelled material.

The racial identification of maize was collected by Dr. Juan Manuel Hernández Casillas, National Leader of Genetic Resources at the National Institute of Forestry, Agriculture and Livestock (INIFAP). The samples were stored in glass jars, sealed and labeled with less than 15% grain moisture, and sent to the INIFAP Central Germplasm Bank located in the Experimental Field Valle de Mexico, for ex situ conservation.

Overall 135 native populations were sampled corresponding to different status and environments where seed samples were collected and five races and 11 interracial crosses were identified (Table 1), of which 67 corresponded to pure races (49.6%) and 68 to interracial crosses (50.4%). The populations were obtained from the three physiographic regions of Nuevo León, Sierra Madre Oriental (42%), Northern Gulf Coastal Plain (30%) and Great Plains of North America (28%), where most race populations (51%) came from the Sierra Madre Oriental, while 49% was obtained in the Great Plains of North America and the Northern Gulf Coastal Plain. Meanwhile, populations of interracial crosses came mostly from the Northern Gulf Coastal Plain (38%) and the Sierra Madre Oriental (34%), while 28% was obtained in the Great Plains of North America.

Table 1 Number of collections of native maize varieties in three physiographic regions of the state of Nuevo León, Mexico. 

Among most abundant races, Ratón (33 collections) and Tuxpeño (13 collections) stand out, with a more or less uniform distribution in the three physiographic regions of Nuevo León, and Cónico Norteño (13 collections), with a restricted distribution in the Sierra Madre Oriental, while Olotillo is also an important race, with three collections, two of them distributed in the Montemorelos and Linares municipalities, in the Northern Gulf Coastal Plain, and one collected in the Dr. Arroyo municipality, Sierra Madre Oriental, in contrast, the Tuxpeño Norteño race has very limited incidence, since only one collection was obtained in the Linares municipality, in the Northern Gulf Coastal Plain (Table 1).

Race populations obtained in this study partially confirm results from previous studies in Nuevo León. One of the pioneer works by ^{Villalobos (1977)}, evaluated 96 maize collections obtained in 1975 in the lower parts of Nuevo León, grouped into early, middle and late varieties, corresponding to three racial groups: Tuxpeño, Vandeño and Nal -Tel, and relative influence of the Cónico and Cónico Norteño races. Another study from ^{Ortega (1985)}, described Ratón and Tuxpeño Norteño as races obtained in Nuevo León.

Ten years later, reports by Cardenas (1995), cited by ^{Turrent and Serratos (2004)} mentioned that the INIFAP germplasm bank stores material from the Tuxpeño, Cónico Norteño, Tabloncillo and Tablilla de Ocho races. ^{Ortega (2007)} reported that the Celaya race is also present in the germplasm bank. Finally ^{Valadez and García (2008)} and ^{Zavala (2010)} summarized that the gene pool of native maize in the state of Nuevo León is composed of five races (Ratón, Tuxpeño, Cónico Norteño, Olotillo and Tuxpeño Norteño) and eight interracial crosses (Tuxpeño x Ratón, Ratón x Tuxpeño, Cónico Norteño x Tuxpeño, Cónico Norteño x Ratón, Olotillo x Tuxpeño, Ratón x Cónico Norteño, Ratón x Olotillo and Olotillo x Ratón).

The results obtained in this study confirm that there are features of genetic erosion in native maize varieties from Nuevo León. This situation is due, among other factors, the lack of permanence of farmers in the field, as in the work by ^{Valadez and García (2008)} and ^{Zavala (2010)} who found that in the rainfed maize production system, there are many abandoned plots; recurrent droughts occur during the crop cycle, because most of the area is operated under rainfed conditions (^{SIAP, SAGARPA, 2011}), where the factors limiting maize production are temperature and water availability in the soil as a result of climate change, as in the neighboring state of Tamaulipas (^{Castro Nava et al, 2011}; Castro Nava et al, 2013). In addition, little research and breeding has been carried out in native maize over the last 20 years (^{Reyes and Cantú, 2006}). Other adverse circumstances are social, such as the advanced age of farmers and the migration of young people to the big cities in the state of Nuevo León and United States of America.

Today, the distribution of native maize in the state of Nuevo León is virtually restricted to rainfed areas in the three physiographic regions in marginalized areas where production systems are low-technified, preventing native maize from being displaced by improved varieties or hybrids of greater profitability. According to ^{Hernández (2000)}, the agroecological characteristics of Mexico mainly topographic and physiographic created specific conditions promoting wide maize differentiation and adaptation to the different ecological niches that were originated.

The presence of large mountain ranges, deserts and other geographic barriers created the conditions for adaptive differentiation of heterogeneous populations. This phenomenon occurs in Nuevo León, where the Sierra Madre Oriental hosts the greatest wealth and abundance of germplasm, while functioning as a protective barrier against genetic erosion; the municipalities where most samples were collected include Galeana, Aramberri and Dr. Arroyo.

In contrast, a different phenomenon occurs in the Great Plains of North America and Northern Gulf Coastal Plain, where the cultivation of native maize has been displaced by pastures for animal feed (^{Valadez and García, 2008}). In these regions, the Olotillo and Tuxpeño Norteño races are materials that tend to disappear from the territory of Nuevo León: the first race has been cultivated for 40 years in elevations between 245 and 303 masl, from where it has been moved to the upper parts of the Sierra Madre Oriental, since one sample was collected in the Dr. Arroyo municipality, at 1 697 masl. The Olotillo race has been reported only in Nayarit (Cárdenas 1995, cited by ^{Turrent and Serratos (2004)}; mainly distributed in the state of Chiapas, where ^{Reif et al. (2004)} have collected it between 300 and 700 masl. Meanwhile, only one sample of the Tuxpeño Norteño race was collected in the Linares municipality at 334 masl. This race has not been previously reported in Nuevo León (Cárdenas 1995, cited by ^{Turrent and Serratos, 2004}; ^{Ortega, 2007}).

Moreover, as mentioned above, just over 50% of the collections made in this work corresponded to interracial crosses, indicating the importance of this genetic resource in Nuevo León, whose presence has not been reported in previous studies, and that somehow involves the generation of new germplasm. According to ^{Dempsey (1996)}, in open-pollinated crops such as maize, landraces can easily interbreed with introduced varieties, and thus may lose some beneficial traits or in some cases, improve their adaptation ability to become local varieties. This process is known as creolization. Interracial crosses of greater incidence in Nuevo León include Ratón x Tuxpeño (19 collections), Tuxpeño x Ratón (17 collections) and Ratón x Olotillo (nine collections); while materials with an intermediate presence were Cónico Norteño x Tuxpeño (five collections) Tuxpeño Norteño x Ratón (five collections), Cónico Norteño x Ratón (four collections) and Olotillo x Tuxpeño (four collections).

Finally Ratón x Cónico Norteño (two collections) and Olotillo x Ratón, Ratón x Tuxpeño Norteño and Tuxpeño x Tuxpeño Norteño, with one collection respectively, are crosses with a limited presence in the three physiographic regions of Nuevo León (Table 1). This trend shows that the germplasm involved in the crosses, has a direct relationship with the most common races in the state, and were distributed near the locations where the races involved in the respective crosses were planted.

Conclusions

Overall 135 seed samples of native maize populations were collected in the rainfed production system in 32 municipalities in the three physiographic regions of the state of Nuevo León (Great Plains of North America, Northern Gulf Coastal Plain and Sierra Madre Oriental). The genetic diversity observed was represented by five pure races (Ratón, Tuxpeño, Cónico Norteño, Olotillo and Tuxpeño Norteño) and eight interracial crosses (Tuxpeño x Ratón, Ratón x Tuxpeño, Cónico Norteño x Tuxpeño, Cónico Norteño x Ratón, Olotillo x Tuxpeño, Ratón x Cónico Norteño, Ratón x Olotillo and Olotillo x Ratón). As some germplasm may be missing further exploration and collection of native maize populations is required in Nuevo León.