Natural Chia Outcrossing

The greatest genetic diversity of Salvia hispanica L. is in Mexico. Currently, its seed derivatives have attracted commercial interest, but little is known about their reproductive system to start and maintain their improvement germplasm. Therefore, we determined the percentage of natural crossing in S. hispanica, in Chapingo, Mexico. We used the flower color as a genetic marker, whereas purple is dominant to white and blue. In 1999, there were two batches of interbreeding, in the first batch, we used seed cultivar of Jalisco and in each row, alternated plants with purple flower and white flower, in the latter, the former cultivar was planted with flower purple, alternating with a wild material with blue flower, collected in Sinaloa. In 2000, the natural crossing was determined based on the percentage of plants with purple flower in F1 progeny cultivar with white flower in the wild type. The material grown in Jalisco had higher outcrossing (22.17%) than the wild material of Sinaloa (1.59%), which could indicate that the mating system of S. hispanica has changed under culture conditions. There intercross reproductive isolation when both genotypes, and therefore should be considered as subspecies or races of S. hispanica. The hybrids are vigorous, as the cultivated parent, and dehiscent fruits like wild progenitor.

Abstract
Mexico has the largest genetic diversity of chia (Salvia hispanica L.). Recently, chia seed derivates have commercial Attracted attention. However, too little is Known About its breeding system to define a genetic improvement strategy and to preserve its germplasm. In this study, the percentage of natural outcrossing of S. hispanica was Determined in Chapingo, Mexico. Flower color was used as the genetic marker, purple flowers Considering dominant over white and blue flowers. Two cross-pollination plots Were Established in 1999. In the first plot, seeds of purple-flowered plants and white-flowered plants of a cultivated genotype from Jalisco Were sown alternately in each row, in the other, the same cultivar Were alternately sown with seeds of purple-flowered cultivated plants and blue- flowered wild plants collected in Sinaloa. Determination of natural outcrossing rate was based on the percentage of purple-flowered plants in the F1 progeny of Both white-flowered plants and cultivation of blue-flowered wild plants. Outcrossing was more frequent in the cultivated genotype (22.17%) than in the wild genotype (1.51%), Which might Indicate That the breeding system was modified under cultivation. Since there was no reproductive isolation, the cultivated and wild genotypes Should be Considered subspecies or races of S. hispanica. The Resulting Were vigorous hybrids, like the cultivated parent, and had Dehiscent fruits, like the wild parent.

INTRODUCTION
Salvia hispanica L. is a plant species that formed an essential part of Mesoamerican culture. For its wide geographic distribution and because knew and used the different ethnic groups, it seems that its cultivation spread from the earliest periods of fitodomesticación, now preserved as marginal crop, the seed is used in the production of soft drinks and nutritious , and its oil in the preparation of lacquers and paints craft (Bukasov, 1963).

The fact that S. hispanica is widely distributed in Mexico combined with its genetic diversity and its food and industrial value, makes it a promising plant resources. Any future program evaluation and use of this kind required to have basic knowledge of their reproductive system.

The description of the floral morphology of S. hispanica was approached by Martinez (1959) and Ramamoorthy (1985) as follows: pedicellate flowers are gathered in groups of six or more, in whorls on the rachis of the inflorescence. The calyx is persistent, pubescent and bilabiate. The corolla is purple or blue monopetalous and bilabiate, the lower lip expands outward and downward, the upper is upward and arches shaped helmet or galea. Fertile stamens are two and are joined by a connective, which is articulated to short filaments that are inserted in the corolla. The ovary is superior, and tetralocular bicarpelar, on the basis of the ovary is nectariferous disc. The style is glabrous, glandular at the base and its stigma has two branches, the longest is the corolla exserted and the shortest is between the anthers. Both anthers and stigma are covered and protected by Galea.

It is not known precisely pollination mechanism S. hispanica. It is assumed that a species is entomophilous allogamous and the color of the petals on the runway as the lower lip of the corolla, the articulation of the stamens to the corolla and the presence of nectar at the base of ovary (Martinez, 1959; Ramamoorthy, 1985). Mann (1959) claims that protandry in the genus Salvia is the mechanism responsible for preventing self-fertilization; meanwhile Haque and Goshal (1981) reported male-sterile plants in S. hispanica requiring certain natural crossing or low seed production in the absence of pollinators. The mechanism of pollination in entomophilous species of the genus Salvia and described by Faegri and Pijl (1979) is as follows: The bees land on the lower lip of the corolla and push into the nectary, this produces the inclination of the anthers that deposit pollen on the back of the insect. Stigma is independent of this mechanism but with some frequency, leans and rubs the back of the insect visitor.

Other authors report that selfing can be present in the genus Salvia. Faegri and Pijl (1979) attributed to the proximity selfing of stigma and anthers; Haque and Goshal (1981) indicate that S. hispanica is self-compatible and that selfing is due to that the flowers are very small and homostílicas and it still occurs in isolated plants seed; Cahill (2004) reported 0.24% of natural cross between a wild and a domesticated population of S. hispanica and although intraspecific hybridization in the wild is rare, controlled crosses between wild and domesticated produce fully fertile offspring.

Several factors influence the mating system in plants. Schoen (1982) mentions that the best predictor floral crossover frequency is the degree of protandry, Wolff et al. (1988) indicate that the presence of male-sterile plants in a predominantly self-pollinating species may increase the frequency of crossovers and Schemske and Lande (1985), floral morphology and the degree of self-compatibility is related to the frequency of crossings. Pollinators seem to be attracted to more intense color flowers (Ennos and Clegg, 1983), the flowers small, with less visual or olfactory signals, are related to self-pollinated (Levin, 1971).

The characteristics of the environment influencing the flow of pollen and therefore in the crossover frequency, is the density (Schemske and Lande, 1985, Wolff et al., 1988) and the abundance of pollinators (Clegg, 1980). The mating system, as other attributes of the genetic system is sensitive to the selection, in Nicotiana rustica, Breese (1959) showed that the morphology and the development of the flower can be modified by selection system and affect mating.

The flower color in S. hispanica can be purple, blue or white. Although there is no information on the inheritance of flower color in this species, some results in other plant species indicate that the flower color is due to the action of a gene pair alelomórficos in the color purple or purple is completely dominant on the target, as in Vicia sativa (Donnelly, 1958), in Phaseolus vulgaris (Miranda, 1969), in Glicine sp. (Weiss, 1949) and in Vigna unguiculata (Sangwan and Lodhi, 1998). Other authors such as Hartwig and Hinson (1963) and Clark and Donnelly (1964), who worked in Glycine sp. and V. sativa, respectively, claim that the genes involved in the inheritance of flower color are located in more than one locus, with different gene action, and that the dominant purple and pink to blue and white dominate the latter. Additionally, Martin and Gerats (1993) have proposed that in addition to the structural genes, which determine the type of dominant pigment in the flower, there are regulatory genes that control the amount and distribution of the pigment, creating shades and hues in the flower between dominant color and white. To estimate the type of pollination under field conditions, it is advisable to use the flower color as a genetic marker, it is a characteristic polymorphic, heritable monogenic or oligogenic type and registration is easy and fast (Bretting and Widrlechner, 1995 ).

Knowledge of the reproductive system of a species allows the choice of an effective strategy for improvement, to select optimal maintenance procedures germplasm and varietal purity and helps to predict their outcome. In order to generate information on the type of reproduction in chia and behavior depending on variety, location and time of year, the objectives of this study were: a) Determine the percentage of natural crossing within and among populations of S. hispanica, using as flower color marker, under the environmental conditions of Chapingo, Mexico b) morphologically characterize the progeny obtained from natural cross between a wild population of Chia Chia and other cultivated.

MATERIALS AND METHODS
The genetic material used was seed plants with purple flower and white flower of a cultivated variety of chia Acatic, Jalisco and donated by Mr. Guillermo Orozco Rose. Although polymorphism has been observed in various characteristics of the variety in culture conditions, were not evaluated segregation ratios in flower color. What is qualitatively observable is the predominance of purple flower on the variety, the white flower was probably the result of a mutation and the selected plant was kept in isolation in reproduction. It was also used chia seed of a wild material, with blue flowers, from the collection of Howard Gentry, University of California, Riverside, and collected in St. Lucia, Sinaloa. In previous evaluations, materials are identified and maintained on flower color characteristic.

The research included two phases: The first was to establish two lots of pollination, in areas far between approximately 600 m, within the Agricultural Experiment Station of the University of Chapingo (CAEUACh), during the season spring-summer 1999.

In both plots pollinated plants were established one month after birth, which germinated and grown in trays in the greenhouse and whose seed was sown on 12 May. The transplant operation was performed on June 12 placing in the first batch, alternately in rows, plants that give purple flowers with plants that bloom white flowers presented, both for the population of Acatic chia. The batch of pollination included eight rows 9 m long, with separation between rows of 0.8 m between plants of 25 cm. The same topological arrangement was established in the second batch with the difference that it is placed alternately in rows, plants have purple flowers were known, the cultivar Acatic Chia, who was known to other blue flowers would, of Chia wild population of Sinaloa. At one end of each batch were placed two boxes of bees to ensure high pollen flow at the time of flowering.

The second phase was carried out in batch CAEUACh SM-17, in the spring-summer 2000. Seed harvested in two batches of pollination previous cycle, was selected from 50 plants of the genotype of white flower and flowering plants of 42 blue Sinaloa. On June 6, from each plant were sown, to “trickle”, three grams of seed per row, they were 8 m long and 0.8 m spacing between them. In order to keep as much of plants that allow crossing record events even in progeny plants with a low crossover frequency of natural, thinning was performed. A flowering, we recorded the total number of plants and number of plants in each row purple flower. For the references to the inheritance of flower color, purple is considered as dominant over blue and white and, therefore, as a marker gene.

To clarify the differences in the characteristics of the flower structure among populations from different geographic origins chia and its possible effect on the frequency of outcrossing, and for the morphological characterization of the progeny obtained from natural crossing between cultivated chia population Acatic, Jalisco and Sinaloa wild population, in a sample of 22 plants from each of the two parents and 22 progeny plants, the following variables were recorded: stem diameter (mm) measured at the first node , length (L) and width (A) of blade (cm), on a sheet located on the fifth knot stem, the ratio L / A sheet of limbo and the number of inflorescences throughout the plant. Two flowers on the middle part of the main raceme of each plant was recorded: the total length of the cup (mm) measured from the base thereof, the color of the flower, the total length of the tube corolla (mm), the width of the lower lip of corolla (mm), the length and width (mm) of the two bracts located in the basal part of the first whorl. A main inflorescence of each plant was measured: length (cm), number of whorls and the number of flowers developed in the first whorl; identified the type of calyx at maturity: open and expelling seed (dehiscent) or closed and which retains the seed (indehiscent). Finally, in two samples of each plant was obtained average weight of 100 seeds (g) using an electronic balance Ohaus brand TS400D model accurately 0.001 g.

The estimation of natural cross-fertilization was done by recording the frequency of dominant phenotypes (purple flower) plants in progenies of recessive (white flower cultivar in Acatic or blue flower in wild material of Sinaloa). Crossing frequencies obtained were grouped into classes to observe the variation between the progenies. The variables measured in the cultivar Acatic in wild material of Sinaloa and their hybrids were analyzed forming two contrasts (contrast 1: Hybrid vs. Parent wild; contrast 2: Hybrid vs. Cultivated parent) using a t test with 0.01 probability (Infante and Zárate Lara, 1986).

RESULTS AND DISCUSSION
Intrapopulation outcrossing
The range of natural crossing in Acatic Chia observed in flower color, ranged from 10.54 to 42.74%, with an overall average of 22.17% (Table 1). 36% of the progeny of plants with white flower had a natural crossing interval of 10.54 to 20%, while most progenies (64%) showed values ??over 20%, with a maximum of 42.74%.

The mean and range of outcrossing obtained indicate that the type of fertilization, the cultivar S. Acatic hispanica is intermediate or mixed type, mating in such selfing occurs within a natural crossing interval ranging from 20 to 80% (Schemske and Lande, 1985) or 10 to 90% (Clegg, 1980 ).

In entomophilous species as S. hispanica certain environmental characteristics as pollinator abundance, fluctuations of insect activity and plant density can influence the flow patterns of pollen and therefore in the frequency of crossing as reported by Clegg (1980); Schemske and Lande (1985) and Wolff et al. (1988). The frequency of natural crossing in the town of Chia Acatic obtained in this research is different from what occurs in their home area. To experimentally established a uniform arrangement between plants, which produce white and purple flowers alternating at equal distance and was provided with a greater abundance of pollinators (two boxes of bees in an area of ??57.6 m2) in Acatic, Jalisco has been estimated a percentage of natural crossing between 10-15%, in planting a “trickle” and their native pollinators. In Acatic used the color of seed coat as a genetic marker, the natural crossing was estimated by the percentage of black seed in the F1 progeny plant with white seed, selected farm plots and seed planted separately.

The presence of variable limits outcrossing in these genotypes indicating that self is present in Salvia hispanica as suggested Haque and Goshal (1981); self-compatible in other species, certain features as the protandry (Mann, 1959; Schoen, 1982) or the presence of male-sterile plants (Wolff et al., 1988), can prevent selfing. In S. hispanica, the presence of male sterile plants determines a higher frequency of outcrossing and low seed production in the absence of pollinators (Haque and Goshal, 1981).

The presence of three phenotypes (purple flower, blue and white) in the F1 crossing of plants producing white flower with purple flower producing plants, both of Acatic cultivar is not consistent with the hypothesis that the color of flower is monogenic, purple being completely dominant over white, as reported in other species Donnelly (1958), Miranda (1969), Weiss (1949) and Sangwan and Lodhi (1998). No contamination could have chia plant pollen type Sinaloa, blue flowers, located on the second batch of pollination as the distance established between the two lots of pollination is within the recommended limits for certain species isolation as entomophilous alfalfa (Medicago sativa) (100-400 m) (SARH, 1980) and also blue flower plants that appeared in the F1 had other morphological characteristics of the variety cultivated Acatic; including indehiscent character, which is recessive. It is suggested, therefore, probably the genes that determine the color of flower in S. hispanica can be located in more than one locus, with different gene action, where the white color of the flower is homozygous recessive expression, as suggested by Hartwig and Hinson (1963) and Clark and Donnelly (1964) for other species. Considering the mixed mating system or intermediate characterizes Acatic Chia cultivar, it is likely that the genotype of some purple flower plants, used in lot of pollination, were heterozygous for one or more loci, and this led to segregation, appearing blue flowering plants in the progeny. Hypothetically it is possible to confirm raised through the crossing of plants homozygous for flower color, selfing of the F1 obtained and assessing the phenotypic frequencies of the F2 progeny.

Interpopulation outcrossing
When people type Sinaloa chia alternately seeded chia population of Acatic, outcrossing limits the type chia Sinaloa were 0 to 7.47%, with an overall average of 1.59% (Table 2). The limits of natural crossing in 78.6% of the progeny were 0.1 to 3% and only four progeny, there was no evidence of cross-fertilization. Cahill (2004) reports, in S. hispanica, 0.24% of natural cross between a wild population collected in Cuescomapa, Guerrero and domesticated from Nicaragua. Crossing values ??are lower than 10% characteristic of populations with a mating system mainly autogamous (Clegg, 1980).

Table 3 shows that plants cultivated chia Acatic have larger flowers and a greater number of whorls and inflorescences with more flowers than wild chia Sinaloa, these features indicate a higher production of pollen in the population Acatic should have generated higher frequency of outcrossing. However in the same frame is recorded that the flowers in the town of Sinaloa are smaller and less intense color, the lower lip of the corolla is narrower, and the smallest difference between the length of the corolla tube and Calyx indicates that the cup more flower wraps, which could have led to preferential pollinator behavior towards flowers Acatic Chia, favoring a higher frequency of self-fertilization in plants Chia of Sinaloa. According Ennos and Clegg (1983), pollinators are attracted to flowers for more intense color and Levin (1971) flowers small, with fewer visual cues, are related to self-pollinated, in S. hispanica, Haque and Goshal (1981) state that self-fertilization occurs in flowers small and homostílicas.

The natural habitat of wild populations of S. hispanica is the pine (Pinus spp.), oak (Quercus spp.) or pine-oak forests of the Sierra Madre Occidental, the Transversal Neovolcanic and the Sierra Madre del Sur. Large orographic and climatic differences that characterize the mountainous landscape have led to geographic and genetic isolation of the populations. According to Stebbins (1957) and Jain (1976) this condition is an environmental factor that favors selfing and Ramamoorthy and Elliott (1998), the geographical fragmentation and inbreeding has led to the diversification and evolution of species of Salvia. In nature there is intraspecific hybridization between wild and cultivated chia because they develop in different geographic regions, whereas controlled crosses between them produce fertile offspring (Cahill, 2004).

Table 4 presents some morphological comparisons between hybrids and their parents. The cup size is the only feature in which no significant differences between the hybrid and its parents, whereas the hybrid shows intermediate values ??different (P ? 0.01) to the parent in the length and width of limbo the sheet, the length of the corolla, the width of the lower lip of the corolla and seed weight. The hybrid not unlike his father cultivated in stem diameter, the flower color, length of inflorescence, the flowers in the first whorl and the number of inflorescences, but the cup is open at maturity (dehiscent) as its wild progenitor.

Ladizinsky (1985) mentions that selfing is a barrier to gene flow between crops and their wild relatives. When hybridization occurs, there is a greater effect on the wild population. Natural selection acts in favor of fruit dehiscence and seed dormancy, ensuring adaptation in natural environments, but also to the cultivated parent characteristics, such as plant vigor and abundant progeny production and fast adaptability to confer growth artificial habitat. The main result of this process may be the formation of breeds with high potential for dispersal and colonization of different natural areas to their original environment and some of these races can be selected and maintained artificially cultivated.

The highest average outcrossing observed in cultivar Acatic Chia, compared with the population of wild chia, could indicate that the mating system of S. hispanica has changed under culture conditions, in Nicotiana rutica, Breese (1959) showed that selection can be changed by morphology and flower development and thus affect the reproductive system. Obtaining chia experimental hybrids, although limited, indicates that parents are subspecies or races of S. hispanica no intraspecific reproductive isolation mechanisms, so that the genetic exchange is possible even between wild and cultivated genotypes of different geographical origin. Therefore, and being allopatric subspecies type, it can be said that their separation is relatively recent evolutionary. To Harlan et al. (1973) the evolution under domestication rarely leads to the formation of new species, the genetic differences between wild and cultivated races are not great, because only a few genes are involved in the domestication process.

Genetic exchange between types of chia could have been more intense in ancient times, when, given the importance of this plant in Mesoamerican culture, wild and cultivated genotypes could coexist throughout the natural range of this species. Even in colonial increased deterioration of germplasm and technological knowledge of chia, is likely in natural or artificial segregating individuals have remained constants hybridizations and have given rise, by selection, to which certain materials grown currently used.

Finally, to increase knowledge about the type of pollination of the species is necessary to develop research on floral structure of populations, the degree of protandry or genotypes in male sterility, insect pollinator behavior and effect under culture conditions.

CONCLUSIONS
Considering environmental and experimental conditions of this study, the findings were:

The genotypes of Salvia hispanica Acatic present an intermediate mating system or mixed, with an average of 22.17% outcrossing and outcrossing limits of 10.54 to 42.74%.

It is likely that the color of flower Acatic cultivar, is controlled by more than one pair of gene action different genes with the complete dominance.

When the wild population of S. hispanica, Sinaloa type was crossed with cultivated Acatic population, the estimated average natural outcrossing 1.51% and a range of 0.0 to 7.47%, indicating a primarily autogamous mating system in the wild.

There is no reproductive isolation between the wild chia and Sinaloa Acatic cultivar, Jalisco, from different geographic origins and degree of domestication, indicating that it is subspecies or races belonging to Salvia hispanica.

The hybrids between the two subspecies are vigorous and high reproductive capacity, as his father farmed, but have dehiscent fruits like wild progenitor.

Revista Chapingo. Horticulture Series
Print ISSN 1027-152X
Rev. Chapingo Chapingo Ser.Hortic vol.14 no.3 Nov. / Dec. 2008
Natural crossing of chia (Salvia hispanica L.)
Natural outcrossing of chia (Salvia hispanica L.)
J. A. Hernandez-Gómez1 *, S. Miranda-Colín2 and A. Peña-Lomelí1

See References, Acknowledgements & Original Article Online