Study site

The experiment was carried out during the spring-summer (S-S) cycles of 2013 and 2014, in the Valley of Mexico Experimental Field (CEVAMEX), belonging to the National Institute of Forestry, Agriculture and Livestock Research (INIFAP), located in Chapingo, State of Mexico, Mexico at 19° 29’ north latitude and 99° 53’ west longitude, at a height of 2250 meters above sea level, with an average annual rainfall of 640 mm and 15 °C average annual temperature (^{García, 1981}).

Establishment of the experiment

Twenty-eight varieties of oats belonging to the CEVAMEX collection were evaluated under temporary conditions on five sowing dates and with a spacing of 07 days between each date. The first sowing date was June 23, the second on June 30, the third on July 7, the fourth on July 14 and the fifth was on July 21, of the S-S 2013 and 2014 cycles. The preparation of the land consisted of a fallow and a harrowing step with the purpose that the ground would remain soft and thus obtain a uniform germination. Seeding was done manually, at a density of 60 kg ha^-1. The presence of the pathogen was allowed to occur naturally, because the environmental conditions where the experiment was carried out are highly favorable for the development of the disease.

We used a completely randomized experimental design with four replications, and an array of treatments in divided plots. The large plots were the five sowing dates and the small plots were constituted by the 28 varieties. Twenty-eight experimental units were established, consisting of four rows with 5 m long × 0.3 m wide, and four repetitions were made of each.

The evaluations were carried out only in the two central rows of each experimental unit and the response variable was the percentage of damage caused by stem rust in each of the oat varieties from the time the disease was disease present until the end of each cycle of the culture. It is worth mentioning that when analyzing the data for a date, the experimental design was completely random; however, when all the dates were analyzed together, a design was chosen in divided plots (^{Montgomery, 2003}).

The incidence and severity data were recorded in all varieties until the end of each cycle; the evaluations were every 8 days in 5 dates and in all varieties.

Variables evaluated

The disease was presented at the beginning of September 2013 and 2014, as of this date the data collection began. In each variety and in each sowing date, the severity was evaluated visually every 8 days until the commercial maturity of the culture using the modified scale of ^{Peterson et al. (1948)}, which consists in taking values of disease severity from 01 to 100%.

On the other hand, the performance of each of the varieties on each date was evaluated. At the end of the 2013 and 2014 crop cycles, 10 stems were harvested at ground level for each variety and in each of the five planting dates. Later the samples were threshed to obtain the grain yield using an analytical scale, and in this way the grain yield of each of the varieties was compared on each date. In addition, the weight of 1 000 seeds without fungicidal treatment was compared against the weight of 1 000 seeds of the same varieties, but treated with the fungicide tebuconazole (Folicur^®, Bayer) at a concentration of 2 mL of commercial product per L of water, sprinkling in total form to the plant during a date of the crop cycle.

Finally, the flowering days were determined by counting the days elapsed from sowing until the time when 50% of the spikelets of each variety reached flowering. While, the days to commercial maturity were obtained by counting the days elapsed from sowing until the moment when the pedicle of the panicle turned yellow.

Statistical analysis

The data obtained from the percentage of severity of all the samplings, in the different sowing dates and in all the varieties during each crop cycle, were adjusted to a model of progress of the epidemic of exponential type with the Curve Expert 2^® program. Likewise, with this model an area under the curve of the progress of the disease (ABCPE) was calculated as an integrative parameter using Scientific Work Place 5^®. A T test was carried out to determine if there were differences in the ABCPE values of the different varieties in the 5 sowing dates and in both cycles. With the statistical package SAS (Version 9.1) an analysis of variance (Anova) and mean comparisons were performed using the minimum significant difference test (DMS) (α= 0.05), by date (completely random) and an analysis of variance and comparisons of means between dates (divided plots).

Variety response to infection

The disease was presented at the beginning of September in both years, time when the first pustules were observed on leaves and stems of the susceptible varieties. On the first date the first symptoms were observed and it was disseminated to the other dates. According to the T test, the ABCPE of the two cycles evaluated did not have significant differences; in such a way that the average obtained from the two cycles was used for the analyzes. The Anova carried out with all the varieties in the five sowing dates showed significant differences (p= 0.05).

When comparing means with DMS, it was observed that the variety with the lowest ABCPE in the five sowing dates was Diamante R-31 (Table 1), which coincided with ^{Jiménez-González (1992)}, who described this variety as one of the most resistant to stem rust for own genes of resistance to different races. Also, the Nodaway, AB177 and Teporaca varieties were also resistant to the infection caused by P. graminis f. sp. avenae, coinciding with what was indicated by ^{Salmeron-Zamora (2001)}, in the case of the Teporaca variety.

The importance of knowing varieties of oats that can serve as a source of resistance to stem rust is key to breeding programs, in addition to the use of resistant varieties is the most used strategy for the control of this disease (^{Mariscal-Amaro et al., 2010}; ^{Leyva-Mir et al., 2013}).

The varieties that were most susceptible and had a higher ABCPE were Guelatao, Tulancingo and Opalo. In the case of the Tulancingo variety, it was deferred with that indicated by ^{Castro-Melendrez and Jiménez-González (1981)}, who considered it to be moderately resistant to stem rust, however, in this investigation it was one of the most susceptible varieties during the two cycles evaluated. Meanwhile, in the case of the Guelatao and Ópalo varieties, our results coincided with those reported by ^{Jimenez-González (1992)}, who reported these varieties as highly susceptible to stem rust.

Comparison between planting dates

According to the Anova, in the five sowing dates (p= 0.05), there were highly significant differences. Likewise, the comparison of means with DMS (Table 2) indicated that planting dates 1 and 2 had the lowest values of ABCPE (severity of stem rust), coinciding with ^{Tovar (1974)}, who assured that there is an effect of different dates of sowing on the yield before the attack of stem rust on varieties of oats.

When analyzing the interactions between each one of the varieties and in all planting dates, it was observed that, in most varieties of oats, dates 1 and 2 could be considered as the best, since they obtained lower ABCPE. It should be noted that some exceptions were found, such as the Gema and Cevamex varieties, where the best sowing date was date 4, as well as the Pampas variety, in which the best sowing date was 3, while in the Raramuri variety was date 5. This shows that there are varieties that are more susceptible to stem rust on certain dates and for this reason the yields of a variety on different dates will depend on the environmental conditions that the disease has to damage the cultivation and decrease the yield, as mentioned by ^{Leyva-Mir et al. (2004)}.

The results generated in this study can be considered to make decisions about the optimal time to carry out the sowing and thus obtain a higher yield of these 28 oat genotypes under rainfed conditions in the High Valleys of Mexico. Similarly, ^{Bobadilla-Meléndez et al. (2013)} verified that planting dates, varieties and densities have a direct influence on the productivity and physical quality of the seed.

Percentage of damage in the last evaluation

Oat varieties evaluated showed an increase in severity from the first to the fourth planting date. Stem rust caused lower percentage of damage in earlier dates compared to later dates that is to say to earlier dates, the oat varieties can escape the attack of rust (Table 3). The above, agrees with the work of ^{Smith (1992)}, who found that one of the best methods of control of stem rust in oats is to choose the right time to sow and avoid late varieties.

The variation in the percentage of damage between varieties shows that each genotype contains genes that give it resistance against the attack of stem rust. For this case we can point out that the Guelatao, Tulancingo, Ópalo, Cuauhtémoc, Texas, Pampas and Perla varieties were the varieties of oats most susceptible to stem rust. While the less susceptible varieties were Diamante R-31, Nodaway, AB177 and Teporaca.

It is important to point out that the majority of oat varieties cultivated in Mexico are susceptible to stem rust, coinciding with ^{Li et al. (2015)}, who mentioned that the majority of the varieties produced in China and worldwide are susceptible to the pathogen and that they vary widely in the level of response to the disease.

Fungicide treatment

The comparison of the average weight of the 5 sowing dates without fungicide and the date with fungicide of the 1 000 seeds indicated that the yield of grain in the presence of stem rust in all oat varieties tends to decrease when the disease is present, because the fungus disturbs the normal source-demand relationships of the plant by sucking the assimilated ones that go to the developing seeds (Table 4). ^{Smith (1992)} found that when P. graminis f. sp. avenae disturbs the epidermis, the plant may dry up, so that the flow of sap does not reach the spike and as a result prematurely matures forming less seed and less weight of the grain.

In the case of the Opalo variety, which is highly susceptible to infection by P. graminis f. sp. avenae, a 52% reduction in yield was found compared to the treatment with fungicide application. This coincided with ^{May et al. (2014)}, who recommended that the application of fungicide for the control of rusts in oats, should only be carried out in highly susceptible varieties. According to ^{Murray (2007)}, the time of application of the fungicides is critical for the success in the control of the disease. Also, the treatments can be applied as treatments to the seed, mixed with fertilizers or sprinkled on the crop.

It should be mentioned that the percentage of loss of yield in some varieties of oats did not coincide with the percentage of susceptibility or damage caused by stem rust, such is the case of the Guelatao variety, which was one of the varieties most susceptible to rust. Stem, although it was not the most lost performance. The above coincided with ^{Epstein et al. (1988)} and ^{Leyva-Mir et al. (2004)}, who mentioned that there are oat genotypes that are highly susceptible to infection by stem rust but that show an acceptable yield because of their high yield potential.

Days to flowering

The varieties with less days to flowering were Diamante R-31 (resistant to rust) and Parámo. This is important, because the earlier a variety is, in addition to choosing an appropriate planting time, it can reduce the risk of loss from stem rust (^{Smith, 1992}). For this reason, the variety Diamante R-31 can be an alternative as a source of resistance if you want to make a cross with some variety of yielding oats but susceptible to stem rust.

The Opalo variety showed to be one of the later varieties in days to flowering and one of the most susceptible to stem rust, which indicates that this type of varieties is more difficult to escape from the attack of this phytopathogenic fungus, even and when the sowing dates are advanced (Figure 1). The flowering days allow us to choose the earliest varieties when the purpose of the production of these varieties is for forage, but when they are required for grain you need to know the days to maturity and thus calculate the planting date in which you can get rid of the attack of stem rust.

Figure 1 Days to flowering of 28 varieties of oats planted under storm in the high valleys of Mexico during the spring-summer 2013 and 2014 cycles. 

Days to maturity

It was observed that the variety Diamante R-31 was one of the varieties most resistant to rust, although its cycle was later compared to the earliest varieties (Figure 2), which behaved as highly susceptible, which indicated that the best way to control this disease is by generating oat varieties that contain horizontal resistance. Similarly, ^{Cornide et al. (1993)} mentioned that horizontal resistance is genetically polygenic and has as its main characteristic the lack of specificity before a series of races, which is why it delays the development of an epidemic or new race.

Figure 2 Days to maturity of 28 varieties of oats planted under rainy season in the high valleys of Mexico during the spring-summer 2013 and 2014 cycles. 

On the other hand, in the susceptible and early varieties, the yields will decrease unless they are sown in times where they manage to escape the attack of the rust, which will be the periods when this disease does not have the adequate conditions to develop and consequently severely damage the crop.