Introduction

The genetic diversity of maize in Mexico is wide, as 59 breeds distributed throughout the country are recognized, from which 1 000 creoles or more are derived in the various localities of the national territory (^{CONABIO, 2012}) material that is protected in different Germplasm Banks in various parts of Mexico by the National Institute of Research, Forestry, Agriculture and Livestock (INIFAP), state universities and in greater numbers by the International Center for Improvement of Maize and Wheat (CIMMYT), with the aim of safeguard the genetic diversity of this species.

The genetic variability of corn constitutes a wealth for the world population and can be the basis for achieving food sovereignty in Mexico, especially in the face of climatic changes (^{Preciado and Montes, 2011}). The evaluation of this diversity is important for genetic improvement programs, due to its potential as a source of new, exotic and favorable characteristics (^{Vigouroux et al., 2008}). There is little information on the usefulness of these breeds and creoles for other regions, so the study of local varieties is important to define their productivity and profitability for forage.

In such a way, that they have a broad knowledge of their productive behavior in order to define a better performance (^{Muñoz-Tlahuiz et al., 2013}), in addition to considering their nutritional quality and effect on animals; thus, producers and technicians will have information and alternatives to supply food and supplements in the production units to counteract nutritional deficiencies in times of greater forage shortage.

Studies carried out by ^{Rivas et al. (2019)} where they compared a local creole from Santa Cruz Ajajalpan, Tecali, Puebla against 12 experimental hybrids and one commercial; observed that the creole outperformed more than half of the improved varieties and was equal to the national average with 21 772 t DM ha^-1, higher than the yields of other improved varieties studied by other researchers such as ^{Borroel et al. (2014)} for the hybrids Caiman, Occorn, AN-423 and Berentsen 302 with 23.83, 20.17, 19.39 and 19.11 t DM ha^-1; respectively, ^{Núñez et al. (2005)}, with 20 t DM ha^-1, ^{Núñez et al. (2007)}, observed values from 21.64 to 16.21 t DM ha^-1 in commercial genotypes, ^{Yescas et al. (2015)}, in the AN-447 hybrid of corn with 14.78 t DM ha^-1 and those of ^{Gaytan-Bautista et al. (2009)} for 22 hybrids from 89 to 11.8 t DM ha^-1.

This shows that some creoles respond better than some improved varieties, offering alternatives for their production, but for many other creoles some information is unknown. In the particular case of San Luis Potosí, there is little information on local or improved corn varieties specifically for the production of forage for ensiling, used by dairy cattle producers in the central, middle and huasteca areas of the state, which improve the systems of production and lower production costs.

Therefore, the objective of this research was to evaluate the productive behavior of 21 genotypes of forage maize of different origin and their adaptation to semi-arid zones in order to offer alternatives to forage maize producers for ruminant feeding, in such a way so that there is availability of genetic material that guarantees a higher dry matter yield to lower production costs.

Materials and methods

The research was carried out at the UASLP Faculty of Agronomy and Veterinary Medicine, located in the Common Palma de la Cruz, Municipality of Soledad de Graciano Sánchez, SLP, at 22° 13' 39.8” north latitude and 100° 50’ 58.3” west longitude and 1 835 masl. The climate is dry temperate with an annual average temperature of 17.1 °C and a rainfall of 362 mm (^{García, 2004}). 21 maize genotypes from different regions and environments were used (Table 1).

Results and discussion

Table 2 shows the comparison of means in PHeight, NLeaves, Diameter, EHeight, FACL, NCorns of 21 corn genotypes from different environments, which showed statistical differences (p< 0.05).

In PHeight the Tlaha2 genotype showed the highest value with 2.5 m, followed by Chalqueño with 2.2 m. and those with the lowest height were the improved Gdelfin and Stiburon varieties with 1.2 m. This shows that the creoles had the highest height and the improved ones the lowest. Results similar to those observed by ^{Parra (1996)} who for 23 genotypes of creole corn observed a height range of 2.45 to 1.69 m. in 1991 and 2.39 to 1.75 in 1992, as well as those observed by ^{Sánchez-Hernández et al. (2011)} for creole corn with 2.44 and 2.16 m in height of a warm humid region. This height is similar, since the creoles that showed a higher height in this investigation come from warm and semi-warm regions.

In NLeaves Tampiqueño1 and Forrajal they showed the highest values with 13.6 and 13.1, respectively, and the lowest value were Mexqui and Tlaha6 with 9 leaves, average. Values slightly lower than those observed by ^{Rivas et al. (2011)} for trilinear hybrids of maize and similar to hybrids and creole ranging from 14.9 to 12.3 leaves plant^-1. In the Chalqueño diameter it showed the highest value with 4.1 cm and the smallest diameter were Ojitalct with 2.1 and Red with 2.0. Values higher than those observed by ^{Rivas et al. (2011)} with 2.71 to 2.62 cm and those of ^{Montemayor, et al. (2006)} with values from 2.12 to 1.97 cm.

The highest EHeight was for Tlaha2 and Papjalb/a with 1.4 m and the lowest values were for Gdelfin and Stiburon 0.5 and 0.4 m, respectively. This shows the variability in terms of the height of the corn on the plant, a character that can be selected in case of requiring varieties with corn located at a lower height.

For the variable FACL, Chalqueño showed the largest area with 776 cm², and the lowest was Stiburon with 519 cm². Character that is observed to be variable between genotypes and that can be selected to improve the FACL in some varieties for a better and greater interception of sunlight, since some researchers have observed that the leaf area index (LAI) is an important biophysical parameter to analyze the amount of photosynthetically active radiation. An increase in the LAI provides increased biomass production; but, due to the self-shading of the plant on the basal leaves, the average photosynthetic rate per unit of leaf area decreases, so special interest is placed on the upper leaves (^{Montemayor et al., 2006}).

Regarding NCorn per plant Papjalb/A, AS948-2, Gdelfin showed the highest number of corns per plant (1.5 corn plant^-1) and the lowest values were for Chalqueño with 0.8 and Rojo and Tlaha1 with 1 corn, respectively. Values higher than the hybrids studied by ^{Rivas et al. (2011)} who observed maximum values of 1.33 corns planta^-1. Characteristic that should matter a lot in the selection or formation of a forage maize, since by showing a greater number they promise a greater amount of grain (corn), which in turn would ensure a greater amount of metabolizable energy for the development of a good silage, since the grain content is one of the main characteristics of corn hybrids associated with the energy value of forage (Allen et al., 1991).

Table 3 shows the statistical significance of the variables total dry matter yield, stem, leaf and corn. With the exception of this last component (YDMC), in the rest of them significant differences were observed between phenotypes (p< 0.05), where Chalqueño and Tampiqueño1 showed the highest values of YDM with 46 246 and 42 947 kg DM ha^-1, respectively in Stiburon change showed the lowest value with 17 300 kg DM ha^-1.

Results greater than those reported by ^{Parra (1996)} with 33 900 to 18 130 kg DM ha^-1, ^{Rivas et al. (2019)} reports 32 800 to 25 000 kg DM ha^-1 for trilinear hybrids, those of ^{Elizondo-Salazar (2011)} with 11 000 kg DM ha^-1 for improved corn and 15 300 kg DM ha^-1 for corn creole, ^{Amador and Boschini (2000)} 15 200 kg of DM ha^-1 and those of ^{Núñez et al. (2005)}, who obtained average dry matter yield results from three maturity stages of 20 000 kg DM ha^-1.

However, values similar to those of ^{Rivas et al. (2011)}, who evaluated 12 experimental hybrids, one commercial released and one creole in San Salvador El Seco, Puebla with the highest fertilizer of 206-69-60 (NPK) and at a density of 100 000 pl ha^-1, under minimal tillage and YDM observed from 47 215 to 34 250 kg DM ha^-1, to those of ^{Sánchez-Hernández, et al. (2011)} with 44 300 kg of DM ha^-1, for a first cycle and in the second cycle 36 600 kg DM ha^-1, and those of ^{Elizondo and Boschini (2011)} with values of 49 203 and up to 38 408 kg DM ha^-1.

In contrast, lower than those of ^{Parra (1996)} for 23 genotypes of creole maize and two commercial varieties, who observed values in the year 1992 of evaluation in a range of 54 100 to 23 000 kg DM ha^-1, with a density of 62 thousand pl ha^-1. An aspect that shows that the yields that can be obtained vary according to the variety, the fertility of the soil, the cutting age and the planting density, among other factors (^{Aldrich and Leng 1974}). If the variety is considered, it can be said that any type of corn can be cultivated for forage, but those that produce higher biomass yields are those tall varieties. For their part, hybrids, being small, generally produce less forage per unit area.

For the YDML, the data showed the highest values for Chalqueño and Tampiqueño1 with 9 781 and 9 528 kg DM ha^-1, respectively, while Gdelfin and Stiburon showed the lowest values with 4 676 and 4 921 kg DM ha^-1, respectively. For YDMS, the data showed that Chalqueño and Mexqui had the highest values with 25 951 and 25 374 kg DMS ha^-1; respectively, however, Stiburon showed the lowest value with 6 883 kg DMS ha^-1, respectively.

If we consider the components of leaves and stems together, it can be seen that the results of this research are much higher than those obtained by ^{Rivas, et al. (2018)} with 5 700 to 4 800 kg DMSL ha^-1 and 19 000 to 13 600 kg DMS ha^-1 and ^{Elizondo and Boschini (2001)} with 2 594 to 1 709 kg DM ha^-1 and 3 327 to 2 802 kg DMSL ha^-1. Although there were no significant differences in the YDMC, the mean of 6 886 kg DMC ha^-1 was higher than that of ^{Elizondo and Boschini (2001)} who observed values of 432 to 667 kg DM of corn ha^-1 and lower than those of ^{Rivas et al. (2018)} with 8 900 to 8 500 kg DMC ha^-1.

Statistical differences in RHP, RSP and RCP are shown in Table 4, where H311plus showed the highest value at 0.29, followed by Stiburon at 0.28 for RHP. Those with the lowest value were Tlaha7 and Mexqui with 0.21 and 0.19, respectively. The highest RSP occurred in Mexqui with 0.67, and the lowest ratio were H311plus with 0.42 and Stiburon with 0.4. On the other hand, in the RCP the highest relationships were obtained in Stiburon and H311plus with 0.31 and 0.28, and those with the lowest ratio were Hueha2b/a with 0.14 and Mexqui with 0.13.

The results obtained are lower than those observed by ^{Rivas et al. (2006)}, who report values between 0.28 and 0.53 as well as those obtained by ^{Reta et al. (2000)}, from 0.45 to 0.5 and to those of ^{Peña et al. (2002)}, which yielded values from 0.32 to 0.4.

The NDF in leaves, stem and corn did not show statistical differences (p> 0.05; Table 5), where the mean was 62.7, 74.9 and 66.7% of NDF for leaves, stem and corn, respectively. Values similar to those observed by ^{Elizondo and Bochini (2001)}, who observed NDF values for leaf from 71.6 to 69.2%, for stem from 68.0 to 65.4% and for corn from 64.9 to 64.1%.

Although the NDF values were by morphological component, it can be seen that the values were lower than those observed by ^{Rivas et al. (2006)}, for silage dry matter that considered the complete plant for the milky state of the corn with 73.4% against 81.4% of NDF in the mass state. Instead, ^{Núñez et al. (2005)} observed lower values with 47 to 60.9% NDF in grain in 1/3 of milk line advance for the complete plant, while with 1/4 of grain milk line, they obtained on average 57.6 to 60.7% of NDF for the entire plant.

The ADF showed significant differences only for stem (p< 0.05; Table 5), where Tlaha6 showed the highest ADF value with 99.6%, while H311plus showed the lowest value with 95.4%. The mean ADF for leaves was 74.3 and for corn it was 95.3%. The results obtained are greater than those obtained by ^{Rivas et al. (2006)} in six corn genotypes harvested in two states of maturity of the corn under minimum tillage, observed that the lowest ADF of 39.1% in the corn harvest in the milky state. In contrast, in the mass state, the average ADF values were 46.42%. On the other hand, ^{Peña et al. (2004)} observed values of 26.4 ADF. In contrast, ^{Elizondo and Boschini (2001)} obtained much lower values with 39.7 to 37.4% for leaf, 44.9 to 42.6% for stem and 34.7 to 25.2% for corn.

Conclusions

The outstanding genotypes in total dry matter yield were two creoles, Chalqueño and Tampiqueño1, which corresponds to the temperate and semi-warm environment. It is concluded that there are creole genotypes from other regions that adapt to the dry climate of semi-arid zones and produce higher YDM and with better morphological components than other creoles in the region and that the recommended improved varieties. The results showed that there are genotypes from different regions and environments, with productive potential for semi-arid areas, so they can be used for production and breeding programs for forage.