Genetic material

Seeds of three varieties of canola were used: Monty, Canomex and Aztecan, and the hybrid Hyola 401 produced in the spring - summer cycle from 2007 to 2012 in Metepec, State of Mexico, with the characteristics indicated in Table 1.

Table 1 Varieties of canola (Brassica napus L.) produced in spring-summer cycle 2007 to 2012 in Metepec, State of Mexico. 

Location of the experiment

Seed quality analyzes were carried out from February to September 2013 at the seed laboratory of the Mexican Institute of Agricultural Research and Training, Aquaculture and Forestry of the State of Mexico (ICAMEX), located in Conjunto SEDAGRO S/N, Colonia Rancho Guadalupe in San Lorenzo, Metepec, State of Mexico.

Experimental development

The seed was obtained from performance tests conducted by the oilseed program of

the National Institute for Forestry, Agriculture and Livestock Research (INIFAP) at the Metepec Experimental Site, Estado de Mexico. In all the years the evaluations were conducted under rainfed conditions during the agricultural cycle spring-summer, the harvest was performed from November 15 to December 15 of each year. Once the seed was harvested, each variety and hybrid was identified with its name and production cycle from 2007 to 2012 and was stored in PVC containers under the natural environmental conditions of the seed store at the test site, with a minimum temperature during this period of 8.3 °C, a maximum of 16.3 °C and an average of 12.8 °C; the mean relative humidity was 62% with a maximum of 78% and a minimum of 40%.

Experimental design and specific characteristics

The experimental design was completely randomized with factorial design (storage time and genotype). Four seed samples of 300 g of each genotype were taken at random, which were used for the following tests: viability percentage (VIA), germination percentage (GER), rate of emergence (IVE), total emergency (EMERT), thousand seed weight (PMS) and volumetric weight (PVOL) (^{ISTA, 2005}).

GER. 400 seeds were placed in petri dishes (50 seeds per box) with moistened filter paper in an SD8900 germinator (Seedburo Inc., USA) at 25 ±1 °C in darkness for 7 days. They were kept with moisture and at the end of the test the total number of seedlings considered as normal was divided by the total number of seeds and the data was expressed as a percentage ^{Raymer (2002)}.

VIA. It was calculated by summing the number of normal plus abnormal seedlings from the germination test which was divided by the total number of seeds sown and the data was expressed as a percentage.

IVE. With the seed, a greenhouse experiment was established to evaluate the emergency speed, using sand as a substrate. The seeds were planted 3 cm deep, under a completely random experimental design with four replicates (25 seeds per replicate). Data recording started the day the first emerged seedling was observed (four days) and completed 14 days later. The emergency speed index (^{Maguire, 1962}) was calculated as follows.

EMERT. At the end of the emergency speed index test, the total number of emerged seedlings was divided by the total number of seeds sown and expressed as a percentage.

PMS. From pure seeds, eight repetitions of 1 000 seeds taken randomly were counted, weighed on an analytical accurate balance Setra SI-410s^® expressing the result in grams.

PVOL. To determine the volumetric weight a beaker with a capacity of 50 ml was used, filled with seeds, flush on top, weighed on an analytical balance Setra SI-410s ®, this procedure was performed in four replicates for each genotype and storage year. The PVOL was expressed in kg hL^-1 for which the corresponding conversion (^{Moreno, 1996}) was performed.

Statistical analysis

With the information obtained in each test an analysis of variance was performed, and means were compared in pairs using multiple range Tukey test (p< 0.05). The analyzes were performed with the SAS package version 9.3 (^{SAS Institute, 2009}). In order to facilitate the interpretation of the results, and considering the different number of years in storage, the analysis was carried out by dividing the genotypes into three groups. In group 1, the results of the varieties Aztecan, Canomex, Monty and Hyloa 401 were included. In the second group, the introduced varieties Monty and Hyola 401 were compared, and in a third group an analysis of the varieties developed in Mexico Aztecan and Canomex.

Comparison between introduced varieties (Hyola 401 and Monty) and local varieties (Aztecan and Canomex)

Germination. The Aztecan variety significantly exceeded (p≤ 0.05) the other varieties; Canomex and Hyola 401 were similar and Monty was significantly lower (Table 2). The seeds stored in 2010 showed the highest germination, showing a reduction of 38.7% in the year 2007. For the interaction, the germination of Canomex showed a reduction of 44.25% from 2010 to 2007, in Aztecan it was 21% and Hyola 401 showed 29.25% reduction, while the greatest reduction was in Monty, all differences were significant (p≤ 0.05) the change in position of Canomex range between years contributed to the significance of the interaction.

Table 2 Test of means of six seed quality characteristics in four canola genotypes stored for two different periods. 

GER= germinación; VIA= viabilidad; IVE= índice de velocidad de emergencia; EMERT= emergencia total; P1000S= peso de mil semillas; PVOL= peso volumétrico; G= genotipo; AA= año de almacenamiento; G*A= interacción genotipo*año de almacenamiento.

Viability. Aztecan seeds viability was greater (p< 0.05), followed by Canomex and Hyola 401, genotypes which did not differ between them, and Monty showed lower viability. For storage, the viability decreased 38.9% from 2010 to 2007 (p< 0.05). Genotype*storage interaction was significant (p< 0.05), all varieties showed its highest average in the seed stored in 2010, except Monty wich showed in both storage periods the lower viability, the other varieties changed position in 2007 (Table 2).

Index of emergency speed. There was no difference between varieties, however, significant differences were detected between years of storage and for genotype per year interaction. The seeds with the longest storage time (five years) reported the lowest averages, significantly different from the seeds stored for two years (2010). For the genotype by storage time interaction, it highlights the variety Monty with the highest average in 2010 (two years of storage), and was the one with lowest average in seeds stored in 2007 (five years of storage); this trend was generalized in all varieties, except for the Hyola 401 hybrid in which the averages are inverted without being significant (Table 2).

Total emergency. The two local materials and Hyola 401, did not show differences between them and were higher (p≤ 0.05) to Monty which showed the lowest average. For the storage time the highest average was shown by the seed of 2010, being significantly different from the one stored in 2007, likewise, in the interaction genotype by storage time a reduction in the EMERT was observed as the storage time increased with changes in the position of varieties (Table 2).

Weight of a thousand seeds. The Canomex and Aztecan showed the highest average and were superior to Hyola 401 (Table 2). The seed Monty showed the smaller weight. For the storage time, the seed harvested in 2007 was higher than 2010. This suggests that this characteristic was not affected by storage time, but by the production environment occurred in each year, with 2007 being the most favorable. In the genotype by storage time interaction, the highest averages were shown in the seeds stored in 2007 in the Aztecan and Canomex varieties, as well as in the hybrid Hyola 401, with a strong fall of Aztecan in the seed harvested in 2010, which contributed to the significance of the interaction (Table 2).

Volumetric weight. In this variable of volumetric weight, the Canomex seeds significantly exceeded the introduced materials and the Aztecan variety was similar to these, with a tendency to exceed them and statistically equal to the superior variety Canomex. For the storage time the highest average was reported in the shortest storage period (2010), although no significant difference was detected between the storage times and the genotype interaction by storage time; the highest average was shown by the Canomex variety in 2007, which was significantly different from the introduced Monty variety, which showed the lowest average in the same year (Table 2).

In general, regarding the consistency in the response of the four varieties through the variables evaluated over the years, the following were observed.

When comparing the introduced materials Hyola 401 and Monty with the varieties Aztecan and Canomex, it was detected that the local varieties exceeded the introduced materials in four of the six evaluated variables; two of physiological quality (GER and VIA) and two of physical quality (PMS and PVOL). The variety that showed the highest average in most of the determined variables was Aztecan which was significantly different to the Monty 2007 variety which showed the lowest values. The Canomex variety and the Hyola 401 hybrid showed intermediate and statistically similar averages in the variables related to the physiological quality: GER, VIA, IVE and EMERT, in the case of the physical variables (PMS and PVOL), the Canomex variety significantly exceeded the Hyola 401 hybrid.

Regarding the storage time, significant differences were detected in the physiological quality variables: GER, VIA, IVE and EMERT, as well as in weight of one thousand seeds corresponding to the physical quality, showing the highest averages in the varieties stored for a shorter time (two years) except for PMS that showed a different behavior, where the highest average was obtained by the seeds with five years of storage, this characteristic was more influenced by the production conditions than those of storage.

Introduced varieties

Germinating. In this comparison, the average of the three storage years Hyola 401 was higher (p≤ 0.05) to Monty, the seed of this variety harvested and stored since 2007 did not germinate (Table 3). Germination decreased according to storage time. When comparing the germination of materials with respect to the storage period, it is shown that although Hyola 401, as referred by ^{Raymer (2002)} decreased along with storage, this was not of the same magnitude as in Mont.

Table 3 Test of means of five seed quality characteristics in two introduced varieties of canola stored for three different periods. 

¹GER= germinación; VIA= viabilidad; IVE= índice de velocidad de emergencia; EMERT= emergencia total; P1000S= peso de mil semillas; G= genotipo; AA= año de almacenamiento; G*A= interacción genotipo*año de almacenamiento.

Viability. Hyola 401 was greater than Monty (p< 0.5) in viability, and the latter stored since 2007 was reduced to 0%, while Hyola 401 decreasing was significant but lower (Table 3). Similar to germination, prolonged storage under natural conditions (suboptimal), showed a differential effect (^{Elias and Copeland, 1994}).

With one year of storage the germination and viability of the canola seed remained at levels allowed for commercialization (>85%), storing it for a longer time significantly affects viability and germination and its use as a seed is not viable. Similar results were recorded by ^{Gusta et al. (2004)}, who determined the seed quality produced at different planting times and noted that seed vigor declines after one year of harvest. Also, ^{Elias and Copeland (1994)} indicated that due to deterioration of the canola seed after 10 months of storage under different conditions, it is necessary to determine the quality each month due to the gradual decline of germination capacity and vigor after 14 months of storage ^{Prakrash (1980)}.

Index of emergency speed. As in the previous features, Hyola 401 outperformed Monty (Table 3). The seed produced and stored in 2012 obtained the highest rate of emergency rate and was higher (p< 0.05) to the seed stored since 2010 and this in turn higher than 2007. Although Hyola 401 seed was less affected in storage, emergency speed loss of 2007 seed, was more than 40% in value compared to that of 2012.

Total emergency. Hyola 401 was higher (p< 0.05) to the variety Monty (Table 3), the seed of 2012 had the highest Emert, higher (p< 0.05) harvested in 2010 and this in turn to 2007. Emert In Monty in 2010 was lower than 2012 at 22.2%, and 2012 to 2007, was reduced 100%. For its part Hyola 401 in 2012 presented the highest average which declined as the storage period advanced. These results are congruent with those reported by ^{Elias and Copeland (1994)} and indicate that the presence of genotype*time interaction, although significant was not very important because of its contribution level to the total phenotypic variance observed for this trait. When comparing the contrasting response of the varieties (Table 3), a low relation was observed between the total seed emergency and the rate of emergency speed. This suggests that the characteristics that control or influence each of these two variables between varieties, do so in different magnitude and consequently gives that low association between them (^{García, 1981}).

Weight of one thousand seeds and volumetric weight. For PMS there were significant differences between genotypes, years of storage and the interaction between them (p≤ 0.01). In contrast, for the volumetric weight no significant differences were observed for the three variation sources, so this feature would not be useful to classify differently each treatment and the results are not shown. In PMS Hyola 401 outperformed the Monty variety. Regarding storage time the highest average was shown by 2010 seeds that were statistically higher (p≤ 0.05) tan 2007 and 2012 seeds in PMS. The Monty variety showed a lower weight in seeds harvested and stored since 2007, but with very little difference to that harvested in 2012 (Table 3).The characteristics of size and weight of canola seeds have been noted as important, as large seed lots, compared to medium and small seeds of the same genotypes, improved planting of seedlings, weight of the area part of the plant, total biomass and yield as well as tolerance to beetles of Phyllotreta genus (González, 1981; 2003; ^{Elliott et al., 2008}).

For the comparison of these materials, in general terms, it was observed that Hyola 401 hybrid significantly exceeded the Monty variety in all variables of physiological quality: germination, viability, emergency rate and total and physical emergency: thousand seeds weight. There was also a trend in favor of the hybrid in the volumetric weight, which can be explained by a higher seed weight or the hybrid vigor inherent in the genotype. Similar results in weight, and even performance and components of the hybrid Hyola 401 were reported by ^{Ortegón (2003)}; ^{Ortegón et al. (2006)} in a study of dates of sowing conducted with a group of varieties and hybrids in the North of Tamaulipas. Also, the size and weight of canola seed were identified as important characteristics by ^{Elliott et al. (2008)}.

Regarding the storage effect, there were significant differences in four variables of physiological quality: GER, VIA, IVE and EMERT, with the highest averages in the shortest storage time; that is, the deterioration process was greater as the storage period increased. However, for the PMS physical variable, the 2010 seed showed the highest averages. This result is probably due to a more favorable production environment for seed production in 2010 and to a reduced effect of storage time on this characteristic. ^{Gusta et al. (2004)}; ^{Franklin (1985)}, mentioned that the production environment has a significant effect on seed vigor, so that localities and production systems that consistently produce high-quality seed should be sought.

Local genotypes. In this case, the most recent seed had two years in storage and was below the values allowed for its use as seed (Table 4).

Table 4 Test of means of five seed quality characteristics in two national canola varieties stored for three different periods. 

GER= germinación; VIA= viabilidad; IVE= índice de velocidad de emergencia; EMERT= emergencia total; P1000S= peso de mil semillas. G= genotipo; AA= año de almacenamiento; G*A= interacción genotipo*año de almacenamiento.

Germination. Aztecan was higher (p> 0.05) to Canomex. Regarding to storage time, the average 2010 seed germination was higher (p> 0.05) than in 2008 and 2007. The effect of the interaction genotype*storage was significant. The germination of Canomex was reduced drastically as the storage time increased, but not that of Aztecan, since its germination of 2010 and 2008 was located in the same statistical group (DMS 0.05), passing to a group of lower germination in 2007. This shows the superiority of Aztecan in this characteristic (Table 4).

Viability. As in germination, Aztecan significantly outperformed Canomex. The highest viability of seed was harvested in 2010, reducing as the storage time increased. Canomex showed the lowest averages in 2008 and 2007, indicating that it is more sensitive to prolonged storage (Table 4).

Emergency speed index. Years and genotypes were determining factors for the expression of emergency speed index, showing highly significant differences (p≤ 0.01), being years or storage time the most important factor. The interaction was significant, detecting the higher averages in the Aztecan and Canomex varieties with the shortest storage time (2010) and statistically different to the years 2008 and 2007.

Total emergency. No differences were observed between genotypes. For years of storage, the total emergency in 2010 was higher (p< 0.05) than in the seeds of 2008 and 2007. From 2010 to 2008 Emert of Canomex decreased by 29% and in 2007 it decreased by 36%, while in the Aztecan variety from 2010 to 2008 it decreased by 36.84% and in 2007, 43.85% showing a change in the position of the varieties from 2008 to 2010, suggesting an important effect of the interaction on this characteristic (Table 4).

Weight of one thousand seeds and volumetric weight. The Aztecan variety was significantly superior to Canomex, and in storage years they are shown from highest to lowest in 2007, 2008 and 2010, respectively. This indicates that the seed weight is independent of the storage period and that the weight was dependent on the variety and the production conditions in each particular year. The genotype per year of storage interaction was significant because there was change of position of genotypes over the years. For the evaluated variable of volumetric weight no statistical difference was detected for the three variation sources.

In general, in the comparison between the two local varieties, significant differences were detected in four variables: GER, VIA, IVE and SMP, with Aztecan superior to Canomex, a similar but not significant trend was detected in the EMERT.

Regarding to storage time of 2, 4 and 5 years, it was shown that the GER, VIA, IVE and EMERT averages were statistically different between 2010 and 2008 and statistically the same those of the years 2008 and 2007 for IVE and EMERT, with a trend of higher averages in 2008. In the thousand seed weight, the highest average was shown in 2007, followed by 2008 and finally by 2010, being statistically different among them. As with the introduced genotypes, neither negative effects of storage time on seed weight were shown in the local genotypes.

Analysis of simple correlations

Total emergence, as an indicator of field behavior and vigor, was significantly correlated with seed viability (0.84), germination (0.84), and emergency rate index (0.86) (Table 5). That is, the physiological characteristics of the seed influence the total emergency, which in turn is an indicator of the response that would be obtained in the field. On the other hand, the negative correlation between PMS and IVE, although not high, suggests that a seed of lower weight emerges in less time or in a more synchronized way than the one of greater weight.

Table 5 Correlation coefficients among seed quality variables evaluated in four canola varieties stored under natural conditions in Metepec, Estado de Mexico. 

Aging of seeds in storage is an inevitable phenomenon, but the level and speed of decline in seed quality depends strongly, in addition to the storage conditions, of the stored species and the initial quality of the seed stored (^{Elias and Copeland, 1994}) as well as the genetic characteristics of seeds. ^{Verma et al. (2003)} studied the seed of two varieties of Brassica campestris that was stored for four years under normal ambient conditions. The percent germination remained above the minimum allowed standards (85%) in the seed of one and two years in storage. Many seed quality characteristics decreased as the storage time increased and the decline was greater in seeds stored for 3 and 4 years. These results are similar to those reported in this research.

Moreover, ^{Nagel et al. (2011)}; Castle (2002) determined the seed longevity 42 accessions of Brassica napus, the comparison showed that germination viability was partly genetically controlled and varied between 42% and 98%. To define the genetic basis of longevity, they subjected a population of double haploids to three aging treatments. They detected QTLs on six different chromosomes, suggesting the complexity of this feature.