Vegetal material

The 41 seeds poblano pepper native populations were collected in 10 municipalities of the Sierra Nevada state of Puebla, Mexico. This area is located west of the state in the area comprised between parallels 19° 02’ and 19° 28’ north latitude and meridians 98° 15’ and 98° 40’ west longitude; three populations of Ancho pepper (A13, A23 and A30) and two Loco pepper in the same region (L43 and L36), a Miahuateco pepper (M48) of the municipality of Santiago Miahuatlán , Puebla and Ancho pepper also collected Fresnillo, Zacatecas (A10); it was also considered the commercial hybrid ancho Doroteo pepper (A49).

Seedling production and localities evaluation

The planting was carried out in styrofoam trays of 200 cavities and as a substrate a mixture of peat moss and forest soil (1:1) was used. Seedling production was conducted in a greenhouse. The towns of evaluation were Chahuac, municipality of Domingo Arenas, Puebla, located 19º 08’ north latitude and 98º 27’ west longitude at an altitude of 2 240 meters and San Lorenzo Chiautzingo, Puebla, located 19º 12’ north latitude and 98º 28’ west longitude, at an altitude of 2 360 meters, both with temperate humid with summer rains (^{INEGI, 2011}). In the first locality he was transplanted to 68 days after sowing (dds) and the second at 77 dds, when the seedlings were 12 cm high and 6 true leaves.

Design and experimental unit

The populations were evaluated in plots of cooperating farmers through an experimental design in randomized complete block design with three replications. The experimental unit consisted of a groove of 4.2 m long and 0.8 m wide, with two floors each 0.35 m.

Variables registered

The five plants were chosen for experimental plot to record 71 morphological and agronomic variables, using the manual descriptors Capsicum (^{IPGRI, 1995}); additionally the variable width (cm) length (cm) and dry fruit weight (g) were measured; fresh fruit yield per hectare and indices of length-width cotyledon leaf, mature leaf, cool and dry fruit was also estimated. Seedling variables were recorded in prior greenhouse transplant and the rest in field grown plants. The five fruits that were measured were obtained from the second crop of five floors full competition.

Statistical analysis

Using SAS program version 9.0 (^{SAS Institute, 2002}), an analysis of variance considering the model completely randomized design for variables seedling stage and combined analysis of variance model design in randomized complete block for the rest held variables; these analyzes allowed to identify the variables with statistically significant differences and then discriminating variables correlated by Pearson correlation analysis. Product of the two previous analyzes were obtained 30 variables that were the basis for the principal component analysis (ACP) and cluster analysis using the method of pairs of groups unweighted arithmetic average (UPGMA, for its acronym). According to the groups formed in the dendrogram of cluster analysis of variance between groups to see if there are statistical differences between them it took place.

Results and discussion

In the combined analysis of variance, 60% of the variables showed statistically significant differences. Similar data were found in pepper water collections in Oaxaca (^{Martínez-Sánchez et al., 2010}), where 66% of the variables showed differences between collections. In the study by ^{Medina et al. (2006)} of 94.2% of the variables showed statistically significant differences because of different pepper six species were evaluated. In this paper, fruit variables were those that showed greater variation, with 82.1% of significance; on the contrary, flower variables were those with fewer significant variables, with only 23.5%. This data is an indicator of high morphological diversity in native populations and that is not being exploited or its use is concentrated in a few variables, such as performance.

Results of analysis of variance and Pearson correlation was obtained a decreased to 30 morphological variables (Table 1), same that were used for multivariate analysis (principal and grouping components).

The principal component analysis showed that the first four components explained 56% of the total variation. The first principal component (PC1) explained 22% of the total variance and was composed of variables regarding the quality of the fruit, as fruit weight and color of the fruit when ripe; CP2, with 18% of the total variation, was influenced by variables on reproductive capacity in the fruit, such as length of the placenta and fruit set; CP3, with 9% of the variation was influenced by variables days to flowering and presence of anthocyanins in plant and fruit; and CP4, with 7% of the variation, consisted of characteristics of plant structure such as height and width of the floor and index length and width of the cotyledon leaf.

The fruit variables were important in this study, since 50% of the variables of principal component analysis were related to this body, followed by variables of the structure of the plant and precocity. The native populations of poblano peppers showed more diversity in variables related to weight and color of the fruit; the latter feature is very important because it is what gives color to the mole main product poblano pepper or mulatto in mature state; Also, the bearing of the plant must be suitable to bear a greater burden of fruit, and finally the precocity for quality fruit market in less time. Regarding the amount of total variation detected in other studies, ^{Kadri et al. (2009)} mention that with the first six principal components obtained 54.3% of the total variation in 48 accessions of pepper of Turkey, where the CP1 was formed by diameter, weight and volume of fruit and CP2 for fruit length and pedicel of the same.

By contrast, studies with a single type of pepper like ^{Moreno et al. (2007)}, who used accessions guajillo from Durango, Zacatecas and Guerrero was found that the first five principal components explained 71.3% of the total variation and the variables that contributed most to the total variation were vegetative and reproductive type. Also, in the work of ^{Martínez-Sánchez et al. (2010)} in Oaxaca pepper water found greater variation in precocity characters and branching density.

Given this, one might consider what ^{Latournerie et al. (2001)} mention, regarding the genetic wealth present in the various regional peppers largely due to the diversity of soil and climatic factors that have been developed, as well as the preservation and management of native seeds for hundreds of years by farmers. In this regard, ^{Medina et al. (2006)} indicate that the existence of genetically diverse germplasm can be included in breeding programs through the selection of characteristics of interest. Meanwhile, ^{González and Bosland (1991)} indicate that the collection and characterization of local populations is necessary to obtain and find additional genetic resources that are not known and that the desirable characteristics can be transferred to commercial cultivars.

Furthermore, ^{Geleta et al. (2005)} suggest that reproductive behavior is determinant in the degree of genetic diversity among cultures; although ^{Lefebvre et al. (1993)} mention that the reproductive behavior of C. annuum L. is very fickle, compared to other self-pollinating species, because most species of Capsicum, among them C. annuum L., are protogynous (^{Pickersgill, 1997}; ^{Djian-Caporalino et al., 2006}) and there is a degree of “excursion” the stigma different for each genotype, which extends the possibility of crosspollination between populations (^{Pickersgill, 1997}). The way of reproduction of these peppers, as well as the empirical selection that have made the producers of the region of the Sierra Nevada de Puebla has given the possibility of finding genetically diverse germplasm, mainly in characteristics of fruit, where you can find variability in weight , size and color.

As a result of principal component analysis Figure 1, where the dispersion of the populations shown based on the first two principal components obtained. In this we can see that most stocks poblano pepper grouped in quadrants I and IV; however, a separation of these populations according to local collection was not obtained, which can be attributed to these populations are located in a small niche.

*Poblano ■ Ancho ● Loco ◊ Miahuateco ○ Híbrido ancho. PDF= weight of fruit; CFM= color fruit when ripe; CUF= ripened fruit; LPL= length of the placenta (variables associated with CP1 and CP2).

Figure 1 Dispersion of poblano pepper populations native to the Sierra Nevada of Puebla, through the first two components. 

The separation of populations poblano of ancho pepper (A13, A23 and A30) and populations loco pepper (L43 and L36) in the region of the Sierra Nevada de Puebla, which were located in Quadrant II, is attributed to one of the most important variables for this study was the color of the fruit when ripe. On the other hand, the collection of ancho de Zacatecas (A10), closest to hybrid (A49) and the population of Miahuateco pepper (M48) were placed in Quadrant III.

However, despite a general grouping in the towns of pepper poblano subgroup formation (Figure 2) it can be seen, encountering this heterogeneity within it. About ^{Votava et al. (2005)} note that in each region processes of natural and artificial selection are performed by each of the farmers and this may be the reason to find such heterogeneity. Also, for the variable fruit weight most poblano pepper populations show higher fruit weight than other types of pepper. For length of the placenta were Loco pepper populations which showed the longest; this due to the elongated shape of the fruit, there likewise increased fruit set for this type of pepper.

Figure 2 Dendrogram of native populations poblano, anchos, locos and a hybrid, generated with the clustering method UPGMA. 

The formation of three subgroups of native populations of poblano pepper and a subset of ancho pepper shown in Figure 2. There existed a group of stocks in accordance with the municipality of collecting possibly related to losses in seedling production by diseases as "damping off" disease caused by a group of fungi consisting of Phytophthora capsici Leo., Fusarium spp., Rhizoctonia spp. and Verticillium spp. (^{Velásquez et al., 2001}), where data found losses of up to 90% (^{Rodríguez et al., 2007}). The germplasm loss caused by this disease require farmers to use other seeds or seedlings available in the region, whereby a flow of germplasm is promoted between locations. In this regard, ^{Rincón and Hernández (2000)} mentioned that the flow of genetic material between farmers and between communities contributes in part to the gradual modification of the genetic diversity of local materials. In Figure 2 can also be distinguished six varieties that are not part of a specific subgroup, reaching the ends of the dendrogram. On the left side is a variety of pepper poblano (P01) and the collection of Miahuatlán (M48). These two varieties, although they were collected in different regions showed similarities to settle at CP1 associated with features of very small fruits and brown.

The data obtained for the four groups formed and populations that are not part of any group are described in Table 2. The analysis of variance between groups (data not shown in table) showed statistically significant differences for all the variables that make up the first four main components: fruit weight (p= 0.0001**), ripe fruit color (p= 0.0001**), ripened fruit (p= 0.0001**), length of the placenta (p= 0.0001**), anthocyanins at the nodes of the plant (p= 0.0066**), days to flowering (p= 0.0030**), anthocyanins spots on the immature fruit (p= 0.0161*), length-width ratio of the cotyledon leaf (p= 0.0002**), the plant width (p= 0.0022**) and plant height (p= 0.0340*).

The results of this research reflect the existence of genetic diversity in native populations of poblano peppers in the study region, whose origin, at least in part, is the team that for years farmers have had on these populations, obtaining desirable characteristics for them and consumers, such as size, color, earliness, length of the placenta, the presence of anthocyanins in fruits and plants, as well as width and length of the plant, yield, among others. This same selection and generate genetically different materials has favored the adaptation of these populations to specific locations and environments.

Proper evaluation and characterization of diversity is important to initiate programs aimed at selection of yielding varieties, rescuing some characters of interest to effectively leverage this plant genetic resource (^{Kadri et al., 2009}), also allowing its conservation, expanding the genetic base of crops and protection for the same (^{Yüzbaşıoğlu et al., 2006}). In this context, ^{Votava et al. (2002} and ²⁰⁰⁵⁾ indicate that the availability and strategic use of genetic diversity inter and intra are the pillars on which rests the future of improvement.