Plant material

Three cultivars of free-pollinated purple corn were evaluated, which were INIA-601, Morado mejorado (MM) and Canteño, whose characteristics are explained in Table 1.

Trial site

The trial was carried out in the Department of Cajamarca in four localities, which are Tartar Chico, Shaullo Chico, Iglesiapampa and Chala, their geographical location is shown in Table 2 (^{Weather Underground, 2020}). The agricultural campaigns in the four localities were carried out from October 2018 to April 2019.

Field management and experimental design

The block design was completely random, with a sowing density of 50 000 plants ha^-1. The experimental unit consisted of five furrows established at 80 cm between furrows, 50 cm between holes, placing two seeds per hole. Each of the furrows was 5.5 m long and the number of experimental units was 4. Fertilization in the four localities was used in the form of N-P-K, its chemical composition is N, P₂O₅ and K₂O, respectively, it was in Tartar chico (165, 45, 65), Shaullo chico (120, 60, 50), Iglesiapampa (120, 60, 50) and in Chala (145, 65, 45) kg ha^-1, respectively. The fertilization was made with island guano twice, the first application was at sowing and the second at ridging.

In all experimental units, pests such as fall armyworm (Spodoptera frugiperda Smith) and earworm (Helicoverpa zea Boddie), were controlled. Evaluations were conducted in the three central furrows and ten plants per treatment were randomly selected. The biometric variables analyzed were: days of female and male flowering, taken from 50% of plants releasing pollen and showing pistils, plant and ear height in meters, prolificacy, defined as the number of ears/plants, ear rot in (%), using a scale from 1 to 6 established by the Maize Program of the International Maize and Wheat Improvement Center (CIMMYT, 2004) and root and stem lodging, both expressed in (%).

The agronomic variables were: field weight, (%) of shelling and area factor. The yield (t ha^-1) was calculated using the following formula: GNY=FW*10EPA*100-%M86*S Where: GNY is grain yield corrected to 14% moisture in t ha^-1; FW is the field weight; %M is the moisture percentage of the grain; (100-%M) is the coefficient of the percentage of dry matter; 86 is the correction coefficient of moisture at 14%; (10/EPA) is the correction factor to transform kg plot^-1 into t ha^-1; EPA is the effective plot area, equal to 4.4 m², and S is the percentage of shelling, equivalent to 0.8.

Determination of anthocyanins

The content of anthocyanins in the bracts and cob of the varieties INIA-601, MM and Canteño was measured in four localities, randomly selecting ten plants per experimental unit following the methodology described by ^{Jing and Giusti (2007}); ^{Gorriti et al. (2009}), and it was characterized by the high-performance liquid chromatography (HPLC) chromatographic method. Samples of 0.3 g of cobs or 0.4 g of bracts, previously dried and ground, were used, placed in a beaker with 100 mL of the hydroacid composed of 850 ml of alcohol and 150 mL of 2% HCl, recording the total weight, it was placed in a magnetic stirrer at 60 °C for two hours at 300 rpm, covering the container with aluminum foil to prevent the evaporation of alcohol.

After weighing, the hydroacid was added until the initial weight was completed, leaving it to stand for 30 min. Five milliliters of sample were extracted and placed in a 100 ml flask, also calibrating with hydroacid and homogenizing, to then measure in a spectrophotometer at a wavelength of 535 nm. To obtain the concentration of anthocyanins equivalent to mg of cyanidin-3-glucoside/g dry weight, the following formula was used: total anthocyaninsmgg=A x MW xDFxVε x l x W Where: A= is the absorbance read at a wavelength of 535 nm; MW= is the molecular weight of cyanidin-3-glucoside, 449.2 g mol^-1; DF= dilution factor; V= volume of the extraction solvent in ml; ε= molar extinction coefficient of 26 900; l= cell path length in cm; and W= sample weight in grams.

Data analysis

Data on yield, morphological and biometric variables and anthocyanins in both cob and bracts were analyzed by locality, calculating the means between the different experiments. The comparison of means was performed using the minimum significant differences (MSD at 0.05), using the InfoStat statistical program, version 2020 (^{Rienzo et al., 2013}).

Yield

Analyses of grain yield, plant and ear height are shown in Table 3 in the four sites evaluated, highly significant differences with p≤ 0.001 are observed for genotypes by environment. Table 4 shows the results of grain yield, it is observed that the best yield was obtained by the cultivar INIA-601, in Chala with 4.38 t ha^-1, in Tartar with 3.35 t ha^-1; however, in Shaullo, the cultivar MM reached 2.2 t ha^-1, a value similar to that obtained by INIA-601 with 2.1 t ha^-1. In Iglesiapampa, the cultivar MM stood out with 1.6 t ha^-1, also followed by the cultivar INIA-601, with 1 t ha^-1.

It should be noted that the cultivar INIA-601 comes from a large population composed of 256 half-siblings, of which 108 come from the variety Morado de Caraz and 148 from Negro de Parubamba. This great genetic variability makes this cultivar very stable in the high Andean areas, whose altitudes range from 2 490 to 3 175 m (^{Pedraza et al., 2017}). It is also known that the experimental cultivar called MM, follows in yield, since this cultivar is an improved synthetic cultivar that comes from INIA-601, while the cultivar Canteño obtained the lowest yield values in the four places of experimentation, probably because this cultivar has an adaptation area in the Lima region at altitudes between 1 800 to 2 500 m (^{Medina-Hoyos et al., 2020}).

In a study conducted in 28 environments and using six cultivars in San Marcos, located in the Department of Cajamarca, yields of 2.77 t ha^-1 in INIA-601, of 2.5 t ha^-1 in MM and of 1.9 t ha^-1 in Canteño were found. In the case of the varieties INIA-601 and MM, lower yield values than those found in the present research work were recorded in some places; while the cultivar Canteño registered a value very similar to that obtained in some tested places (^{Medina-Hoyos et al., 2020}).

The results of plant height indicate that the cultivar MM was superior in all cases, except in Tartar. It is observed that, in Iglesiapampa and Shaullo, the values varied from 1.7 m to 1.95 m, that is, values lower than 2 m, while in Tartar and Chala, it was observed that plant heights were higher than 2 m, with INIA-601 standing out in Tartar with 2.77 m and in Chala, the cultivar MM with 2.32 m, in addition, in all cases the cultivar Canteño showed the lowest plant heights, from 1.67 to a maximum of 2.12 m. It should be noted that plant height is important to estimate or know this characteristic of each variety and they have the disadvantage that if they grow more than average they flake.

A correlation of the values of ear height with plant height is observed in the different varieties, the values obtained in the locality of Tartar were for INIA-601 (2.77 and 1.6 m), MM (2.7 and 1.58 M) and Canteño (2.12 and 1.3 m), followed by a lower plant and ear height in the locality of Chala, for INIA-601 (2.14 and 1.03 m), MM (2.32 and 1.25 m) and Canteño (2.12 and 1.13 m), respectively. On the other hand, it is also observed that, in the localities of Iglesiapampa and Shaullo, plant height values ranged from 1.6 to 2 m and ear height values ranged from 0.8 to 1 m.

The evaluated results of male and female flowering and root and stem lodging using the cultivars INIA-601, MM and Canteño show highly significant differences with p≤ 0.001 for genotypes by environment (Table 5). In Iglesiapampa, the variety MM was the earliest, it reached flowering (male and female) in the shortest time (Table 6) with 94.8 and 96.75 days, followed by the cultivar INIA-601 in Shaullo with 92.5 and 104 days, respectively. It should be noted that the three cultivars in Iglesiapampa, Tartar and Shaullo obtained small variations in both female and male flowering between 92.5 and 111.8 days; however, in Chala, the three cultivars had late flowering periods, their values varied between 143 to 155 days, but with a higher yield.

The values of flowering days are important to know that the earlier the flowering, the lower its yield, compared to that variety whose flowering is late, as shown by the present research work, which shows a late flowering in the locality of Chala in the three varieties and the highest yield compared to the other localities, with the variety INIA-601 standing out with a female flowering of 155 days, male flowering of 150 days and with the highest yield of 4.4 t ha^-1. The lowest root lodging values occurred in Iglesiapampa using the three cultivars with the same statistical value of 0.71%, followed by those obtained in Shaullo and Chala with values in a range of 1.28-1.61%, while in Tartar, the highest values, from 3.28 to 4.08%, were obtained.

The stem lodging values had a similar pattern, the lowest values for the three cultivars occurred in Iglesiapampa, values similar to those obtained in Chala with a range of 0.71-1.52%, while in Tartar and Shaullo, higher values, from 3.26 to 5.64%, were obtained. As can be seen in the results, the locality of Tartar had the highest values of both root and stem lodging, since the soil probably had many nutrients and therefore the plant grew well above the average compared to the other three places and this caused it to fall.

It should be noted that there are two types of lodgings in purple corn, root lodging and stem lodging. Root lodging occurs when the stem falls more than 30 degrees from the vertical and stem lodging occurs when the stem breaks below the ear. Any type of lodging makes mechanical harvest difficult and leads to losses from 5 to 25% (^{García and Watson, 2003}).

The evaluated results of proliferation and ear rot using the cultivars INIA-601, MM and Canteño show highly significant differences with p≤ 0.001 for genotypes by environment (Table 7). The prolificacy, which is the number of ears plants^-1 (Table 8), shows that if the value exceeds 1, it means that each plant has one or more ears, as is the case of INIA-601 in Shaullo, whose value is 1.3, which indicates that some plants have more than one ear, for the case of MM in the locality of Chala, it is 1.13 and for Canteño in Shaullo, it is 1.54, meaning that out of every three plants, two of them have two ears.

For the variety INIA-601, in all localities all plants had at least one ear; however, for the variety MM in the localities of Iglesiapampa, Tartar and Shaullo, there were plants that did not have ears and therefore the values are 0.77, 0.88 and 0.99, respectively, that is, the values are less than 1. The evaluation of ear rot shows that the cultivar INIA-601 obtained the lowest value with 5.9% in Chala, reaching up to 11% in Iglesiapampa. In the four places, the cultivar MM obtained statistically similar values, between 9.1 and 10.2%, while the cultivar Canteño obtained the highest rot values, reaching up to 18.8% in Shaullo.

This parameter is important in the agronomic study, since ear rot is generated by fungi, which not only reduces the quality of the product but also brings consequences to human and animal health, mainly due to the production of mycotoxins. This damage depends on the genotype, environmental factors such as the varied temperature in the crop cycle, soil moisture and relative humidity where the trial is carried out, in addition, the damage caused by birds is greater because it promotes the presence of fungi. Another aspect is the damage caused by the insect Euxesta sp., which attacks the ear from the formation of the pistils and continues during the development of the grain, producing a humid environment at the apex of the ear that facilitates the proliferation of fungi and therefore, ear rot (^{Medina-Hoyos et al., 2020}).

The results show that the anthocyanin content evaluated in cob and bracts using the three cultivars INIA-601, MM and Canteño in four places shows highly significant differences with p≤ 0.001 for genotypes by environment (Table 9). Analyses of the anthocyanin content in cob were higher than the anthocyanin content in bracts, as can be seen in Table 10. The anthocyanin content in cob ranged from 4.6 to 7.9 mg g^-1 of dry matter in the three cultivars and four places, with INIA-601 (7.9 mg g^-1), MM (7.2 mg g^-1) and Canteño (5.5 mg g^-1) standing out in Chala, followed by Shaullo, the value of INIA-601 was also 7.9 mg g^-1, followed by MM with 6.9 mg g^-1 and Canteño with 5.4 mg g^-1.

In the anthocyanin content in bracts by genotype in the four places, there was little variation with significant differences between genotypes by environment, with the cultivar INIA-601 standing out with a range of 3.7-4.9 mg g^-1, MM varied from 1.6 to 2.3 mg g^-1, while the variety Canteño showed very small values of ≤0.2 mg g^-1. As shown, the highest values of the amount of anthocyanins in the cob and bracts were obtained in Chala in the three varieties, likewise, their grain yields were also the highest, with the variety INIA-601 standing out with a yield of 4.4 t ha^-1, amount of anthocyanins in cob of 7.9 mg g^-1 and in bracts of 4.5 mg g^-1, with a good plant height of 2.14 m and ear height of 1.03 m, suggesting to be a potential variety for the import and export of purple corn.

Purple corn produces anthocyanins throughout the plant, especially in the cob and bracts, their values vary significantly (^{Fernandez-Aulis et al., 2019}). The amount of anthocyanins depends on the extraction method, which must be validated, and on the factors involved such as solvent, temperature, pH, stirring time, solvent concentration and the mass-solvent ratio used in both classical and emerging technologies, until now there is no uniformity to obtain anthocyanins. If a cultivar has a good grain yield and a good amount and concentration of anthocyanins, it will generate profitable profits for the farmer.

It is known that for the cultivar INIA-601, it is estimated that for every 5 000 kg ha^-1 of ear, 500 kg of cob and 200 kg of dried and chopped bracts can be produced, with an anthocyanin content of 6.12 and 3.18%, respectively, which is the way these products are marketed (^{Medina-Hoyos, 2020}). The price of dried purple grain is $2.00 kg^-1 and the kg of cob or dried bracts is $20.00 and therefore, in 700 kg of cob and dried and chopped bracts, an income of $14 000.00 is generated. This means that producing 2 500 kg of grain at 14% moisture and 700 kg of cob and dried and chopped bracts in one hectare, the farmer could obtain a gross income of $19 000.00. Subtracting the cost of production that approximates $8 000.00 ha^-1, the net profit would reach $11 000.00 ha^-1 (^{MINAGRI, 2017}).

As can be seen, maintaining fixed costs and selling the cob and bracts with a high concentration of anthocyanins and productivity, greater profitability would be obtained in the cultivars of purple corn. In a study published by ^{Medina-Hoyos et al. (2020}), using six cultivars in 28 locations in the Department of Cajamarca, province of San Marcos, the anthocyanin content in cob ranged from 4.66 to 6.12 mg g^-1 and the anthocyanin content in bracts between 0.63 to 3.18 mg g^-1, values lower than those reported in the present research work. It should be noted that the amount of anthocyanins in the post-harvest stage decreases as the time of permanence of purple corn in the field increases, as reported by Medina-Hoyos et al. (2020), a factor that must be incorporated into the post-harvest protocols in order to obtain high economic results in the sowing of purple corn and its anthocyanin content in both cob and bracts.