Introduction

In Aguascalientes, the family dairy production system (herds smaller than 30 cows in production) is closely related to rainfed agriculture practiced in the state, mainly through the cultivation of forage maize, which led to the sowing of marginal areas, with limited edaphoclimatic conditions. Maize crops supports milk production by producing forage for backyard cattle feeding (^{Luna-Flores and Gaytán-Bautista, 2001}; ^{Carranza et al., 2007}). However, in this area intermittent or terminal drought and the poor level of soil fertility, limit the production of this grass under the traditional production system (^{Osuna et al., 2012}). This causes an annual deficit of more than 350 000 t of fodder. The most critical deficiency occurs in the spring period, given the lack of rain and a shortage of irrigation water (^{Peña et al., 2012}). For this reason, it is necessary to evaluate the capacity of dry matter production and adaptation to the environmental conditions of the region of other forage species, as well as the implementation of agronomic practices that are efficient to make better use of rainwater.

In this temperate semi-arid region of central México, many species with forage potential have not been used yet; such is the case of brown midrib sorghum (Sorghum bicolor L. Moench). Knowing the lack of fodder that exists in the state of Aguascalientes, especially due to water scarcity, it is essential for new fodder crops to be adapted to this region. Sorghum is a crop that requires less water than maize, so it has potential as a forage plant that can be grown in this region, and its nutritional value is equal to or slightly higher in comparison to maize, and with a superior productive response (^{Bolaños-Aguilar et al., 2012}).

Despite the foregoing, and particularly under rainfed conditions, the species itself can not guarantee high yields, so it is necessary to accompany its sowing with suitable in situ rainwater collection practices, that involves techniques that take advantage of rainfall (because it increases the amount of water available for plants), follows practices that help conserve soil, with consequent benefits (^{Martínez and Jasso, 2004}; ^{Osuna et al., 2007}; Ventura and Acosta, 2008).

The use of the corrugation system with “Aqueel” and “furrow diking” are two practices that consist of the construction of micro-reservoirs on the land surface and transverse land borders in the middle of the grooves, respectively, to store the rainwater and to eliminate or reduce water losses due to runoff. The use of these practices has helped to increase crops yields under rainfed conditions (^{Osuna et al., 2007}; ^{Padilla et al., 2008}; ^{Osuna et al., 2015}).

On the other hand, the management of the sowing density is one of the most recommended agricultural practices to achieve an increase in crops yield, because with an appropriate number of plants per unit area, a better use of the water and nutritional resources is achieved. Increasing density reduces biomass and yield per plant; however, the production of biomass and seed yield per unit area are higher (^{Soltero et al., 2010}; ^{Bolaños-Aguilar et al., 2013}).

In order to increase yield per unit area, in addition to the traditional simple groove method, others have been adapted such as planting in double and triple row. Such methods include seeding, double rows in the first, 20 cm within each pair and 80 cm between pairs of rows, and in the second to set beds of 1.6 m with three rows at 40 cm between lines (^{Rodríguez et al., 1994}; ^{Osuna et al., 2015}). Thus, sorghum can be planted with wheat seeder covering three holes and leaving two for double row or covering two and leaving three free for planting at triple row (^{Soltero, 1992}; ^{Rodríguez et al., 1994}).

However, there is a very wide range in population density that is used when planting in double or triple row, which ranges from 200 thousand to 400 thousand plants per hectare. This wide range is because to date there is no enough information both regarding the method and the optimal density (^{Rodríguez et al., 1994}). The population density and the spatial distribution of plants are factors associated with the effective management of this culture to express its full yield potential (Maroni et al., 2003; ^{Bolaños-Aguilar et al., 2013}).

Based on this background, this research was carried out in 2014 with the objective of improving yields and forage quality of maize and sorghum under rainfed conditions with 1.6 m wide beds with four and six rows sowing to evaluate the advantages and disadvantages compared to traditional single-row sowing.

Materials and methods

In 2014, a paper about maize and sorghum was carried out at the Sitio Experimental Sandovales, Aguascalientes, México, located at 21° 53’ 09” north latitude and 102° 04’ 14” west longitude, at a height of 2 100 masl, where an average of 300 mm of rainfall is recorded in the growing cycle, an average temperature of 16.3 °C and a crop cycle of 110 days (late June to mid October); the soil is ±0.3 m deep, with less than 1% organic matter, clay-sandy loam texture, 2% slope and pH of 6.6. Under these conditions, two semi-commercial lots were planted aiming to evaluate three sowing systems and their combined effect on yield and quality of maize fodder (Cafime) and “brown midrib sorghum at high planting densities.

The sowing methods evaluated were: a) rows at 76 cm in single row for maize and sorghum, b) wide beds of 1.6 m with four rows for maize and c) wide beds of 1.6 m with six rows for sorghum. In the case of wide bed sowing, it was established according to the curves at the level, while the traditional sowing was established in straight rows made transversely to the main slope of the land. The experimental unit was of 15 rows to 76 cm, and 8 beds of 1.6 m wide per crop, and 160 m of length for both sowing methods. Plants density per hectare evaluated in maize and sorghum crops were: maize 90 000 plants ha^-1 for wide bed planting of 1.6 meters wide with four rows at 30 cm apart between rows and 30 cm between plants; sorghum 372 000 plants ha^-1 for wide bed sowing of 1.6 m with six rows, its distribution was 10 to 12 seeds per linear meter. Maize and sorghum with 45 000 and 132 000 plants ha^-1, respectively, for single row sowing.

The ground was barked and crawled before sowing, the previous crop was rainfed bean. The sowing took place on June 28 in damp soil. The maize was planted with a versatile precision mechanical seed drill designed for sowing in wide beds of 1.6 m wide with four rows, coupled to a system of water harvesting (Aqueel roller that prints micro-basins on the wide bed for the in situ collection of water and reduced soil erosion), which is located on the back of the drill. Sowing sorghum was carried out with a mechanical drill for forage crops intended for planting in wide beds of 1.6 m wide with six rows, coupled to a in situ water harvesting system (Aqueel roller).

Both equipments were designed by the Mechanization Program of the Campo Experimental Pabellón, INIFAP (^{Rojas et al., 2013}). In addition, in the wide beds sowing it was implemented the furrow diking system for both cultures; this practice consists of raising a 20 cm high ground edge at the sides of the seedbed at regular distances to store water and reduce soil erosion. Traditional maize sowing was performed with a conventional mechanical seed drill owned by the cooperating producer. Traditional sorghum was planted with the INIFAP mechanical sowing machine for fodder crops, which was adjusted to perform furrow sowing at 0.76 m. Soil was fertilized with the 40-40-30 NPK formula applied in the weeding; this practice took place 20 days after sowing; insecticides were applied to control some foliage pests and herbicide for weed control.

The measured variables in climate, soil and plant were:

Weather: In the year the research, the maximum and minimum temperatures, evaporation and rainfall were recorded daily. This information was obtained from the automated meteorological station of Sandovales, located at a distance of 300 m from the experimental unit. The accumulation of heat units (UC) was calculated by the residual method (^{Morales and Escalante, 2006}).

Where: Tmax= maximum daily temperature (°C); Tmin = daily minimum temperature (°C); TB= base temperature or threshold (10 and 15 °C for maize and sorghum, respectively) (^{Ruíz et al., 1998}). Evapotranspiration in both crops under different sowing methods was estimated by the FAO method (^{Allen et al., 1998}).

Soil: prior to sowing, soil samples were taken between 0 and 30 cm depth, in strata of 10 cm (0-10, 10-20 and 20-30 cm), in which texture, organic matter (MO), pH and electrical conductivity (CE) of the production unit were determined. The following procedures were used: texture (Bouyoucos hydrometer); CE in extract; MO (Walkley et al., 1982), and pH in a 2.5:1 water/soil ratio (^{Page et al., 1982}).

Plant: growth stages recorded for aize and sorghum were: days to emergence (EC₁), vegetative development (EC₂), flowering beginning (EC₃) and physiological maturity (EC₄). Harvest was carried out in all treatments at 97 dds, when the maize kernels showed a milky-dough-like appearance and dough-like-grain in sorghum. The cut height was 5 cm from the ground level in all cases. The cut forage was weighed and a 0.5 kg subsample per harvested plot and treatment was taken, which was taken to the laboratory to be dried in a forced air oven at 60 °C until constant weight was reached, to obtain the percentage of dry matter and transform the results on a dry basis (^{Reta et al., 2007}).

The quality of the forages was evaluated through a bromatological analysis. This implied the determination of crude protein percentages (PC), acid detergent fiber (FDA) and neutral detergent fiber (FDN). The near-infrared absorption technique (NIR Systems, Inc., Silver Spring, MD 20904, USA) was used.

To determine the yield of dry matter, ten plots 2 m wide and 5 m long were randomly taken into each experimental plot. The information of the quantified characteristics in the cultures was analyzed based on a completely random design, with ten replicates per crop. For data analysis package Statistical Analysis Systems, version 8 (SAS, 2009) was used and when significance was detected between treatments, the least significant difference test (DMS₀₅) was applied.

Results and discussion

Figure 1 shows the maximum (Tmax) and minimum (Tmin) temperature data; in the latter is observed that the decadal average during the development of maize and sorghum fluctuated between 24 and 26 °C, for Tmax and between 11 and 14 °C, for Tmin. It should be noted that both Tmax and Tmin remained constant throughout the crop development. Seasonal precipitation was 575.4 mm, 81% (466.3 mm) occurred during the crop development (118 mm rained in a single event in 1 day, which represented 25% of the total precipitation). Of the 466.3 mm, 76% (349.7 mm) occurred during the vegetative stage and 24% (116.6 mm) during the reproductive stage of maize and sorghum; this indicates an erratic distribution for crop needs.

Figure 1 Maximum (Tmax.) and minimum temperatura (mean and decennial Tmin) and precipitation (decennial sum) during rainfed maize and sorghum crop cycle. Sandovales, Aguascalientes. 2014.  

Regarding to the phenological stages of both evaluated crops, they were similar: in CAFIME maize, emergence (EC₁) occurred at 8 DDS, the vegetative period (EC₂) from 8 to 55 days DDS, the flowering onset (EC₃) at 62 days DDS and physiological maturity (EC₄) at 97 DDS, in both sowing methods. For sorghum, EC₁ occurred at 9 DDS, EC₂ vegetative period from 10 to 80 days, EC₃ flowering beginning at 85 days and EC₄ physiological maturity at 100 DDS, both in traditional seeding and in wide beds with four and six rows on the edge, respectively.

UC accumulation, related to the phenology of maize and sorghum, are shown in Figure 2. The thermal requirement of maize to achieve each stage was 101 UC at EC₁ (emergency), 664 UC to EC₂ (vegetative period), 815 UC to EC₃ and 980 UC to EC₄ (milky-doughy grain). For sorghum, in emergence, vegetative period, flowering beginning and cutting stage, UC were 46.2, 322, 422 and 505, respectively. ^{Medina et al. (2006)} reported that the El Llano region of Aguascalientes is above 2 000 masl, and these regions are not recognized as viable for the production of sorghum for grain or forage due to environmental conditions, Table 1 shows that the availability of heat using a base temperature of 15 °C, in the summer only 714 UC are available, which differs in more than 2000 UC to the sorghum regions par excellence such as Tamaulipas and Sinaloa.

Tmáx.= temperatura máxima; Tmín.= temperatura mínima; OTérm.= oscilación térmica; UC= unidades calor para maíz y sorgo.

Table 1 Heat units available for maize and sorghum cultivation in Sandovales, Aguascalientes with 10 and 15 °C base temperature, 2014. 

Figure 2 Heat and evapotranspiration units accumulated in the maize and sorghum agrosystem. Sandovales, Aguascalientes. Summer 2014.  

Considering a growing period from June to September (120 days), which is when there is more than 70% of UC in the year (714 UC), it could be thought that for a crop like sorghum that requires higher temperatures, the indicated period would be the most suitable for its cultivation under rainfed conditions. Martín del Campo et al. (1987) mentioned that, even when the heat availability is reduced, when observing the average thermal oscillation trough the year, it turns out to be higher than the one in the regions indicated above.

This characteristic of the region determines that high temperatures are present during the day and cooler temperaturas during the night (lower), which, for a C-4 photosynthesis type such as sorghum, is extremely important for the production of dry matter, since temperatures are within the optimum during the day for the accomplishment of photosynthesis, and on the other hand the low temperatures of the night reduce the respiratory rates and consequently the dry matter production is more efficient (^{Keating and Carberry, 1993}).

Another environmental factor that greatly influences the dry matter production of a crop is potential evapotranspiration. Figure 2 shows the relationship that existed between accumulated evapotranspiration (ETc) of maize and sorghum and its phenology. From sowing to emergence it was observed that ETc was 36 mm for both crops. Because at this stage of plant development it is very limited, it is assumed that most of ETc is direct evaporation of the soil. For the emergence-flowering period, ETc was 257 and 353 mm for maize and sorghum. The ETc accumulated throughout the crop cycle was 403 mm (^{Morales and Escalante, 2006}).

Physical and chemical characteristics

The texture of the soil is sandy loam and clay-sandy, with a pH of 6.6 to 7.6, the lowest value corresponded to the layer 0-10 cm. CIC ranged from 2.2 to 4.6 Cmol⁽⁺⁾ kg^-1, the MO from 0.5 to 1.3% N total from 0.1 to 0.2% (Table 2). Therefore, the soil fertility in these agrosystems is low. In general, it is a soil with strong compaction problems throughout the profile, since the variation range of the apparent density of the studied soil is high and ranged from 1.43 to 1.54 Mg m^-3, values tended to increase with depth soil due to the reduction of biological activity in the horizon A. They have low organic matter content and limited ability to retain moisture, their values are typical of arid soils (Salazar-Sosa et al., 2010).

DA= densidad aparente; Pt= porosidad total; CIC= capacidad de intercambio catiónico; MO= materia orgánica

Table 2 Physical and chemical characteristics of the soil in the agrosystem where the alternative rainfed crops were evaluated. Sandovales, Aguacalientes. 2014. 

The increase in the total amount of water in the soil contributed to the crop during the whole cycle, by the combination of sowing in wide beds in the edge and the microcaptation in situ (with Aqueel system and furrow diking) averaged 32% extra in relation to the traditional sowing system (ST) with straight rows in the direction of the slope and without rainwater harvesting.

Figure 3 shows the retained lamina in the soil profile at 0.3 m depth during the development of the two crops in their different phenological stages, with and without water collection. It is shown in this figure that maize and sorghum planted in wide beds at the edge with four and six-rows (SC-MS) had a larger retained water amount in all growth stages than the traditional system with the same crops and without water catchment.

Figure 3 Water lamina retained in the profile (0.3 m) during the maize and sorghum development in its different phenological stages under two sowing systems. Sandovales, Aguascalientes. 2014.  

The average values of retained lamina in both planting systems for different growth stages of maize and sorghum were EC₁ [SC-MS 85.08 mm vs ST-MS 78.31 mm], EC₂ [SC-MS 82.72 mm vs ST-MS 63.8 mm]; EC₃ [SC-MS 90.29 mm vs ST-MS 70.49 mm] and EC₄ [SC-MS 89.21 mm vs ST-MS 55.43 mm], respectively; it is observed that in all stages of crop growth the highest values of retained lamina were recorded in SC-MS, which is attributed to the combined action of the corrugation practices with aqueel and furrow diking and the wide beds at the edge that produced in situ storage almost as total as rainwater and eliminated the runoff and also decreased the water output to the soil.

Dry matter yield, forage quality and rainwater efficiency

Significant statistical differences were found in dry matter yield and in forage quality between treatments. Table 3 shows that maize yield in the two sowing methods was statistically similar (3.49 and 4.2 t ha^-1), but less than sorghum under wide bed planting in 1.6 m with six rows. The maximum yield for both maize and sorghum was obtained under wide beds sowing of 1.6 m width with four and six rows, respectively. On the other hand, in the two topological distributions, sorghum surpassed maize. ^{Reta et al. (2007)} and ^{Bolaños-Aguilar et al. (2013)} found that by reducing the distance between rows in forage maize sowing by 38 cm, the yield of dry matter increased by 15.5%, whereas in brown midrib sorghum, by reducing the distance between lines to 20 cm, the dry matter yield increased 68% on average, compared to the conventional sowing to 76 cm, which agrees with the results obtained in this research.

Ɨ Siembra de maíz y sorgo a 76 cm entre surcos; PC= proteína cruda; FDN= fibra detergente neutro; FDA= fibra detergente ácido; EUALL= eficiencia en el uso de agua de lluvia.

Table 3 Dry matter yield, maize and sorghum forage quality characteristics and efficiency in wáter use under different sowing systems. Sandovales, Aguascalientes. 2014. 

Regarding to fodder quality, the content of crude protein (PC) of sorghum was higher (p< 0.05) than maize, while in FDA and FDN it was the same (p>0.05) or lower (p< 0.05), according to the evaluation. The quality of evaluated sorghum forage in the two sowing methods is considered acceptable, mainly due to its PC content and low values of FDA and FDN (Table 3).

Regarding to the efficiency of rainwater use (EUA_LL), in terms of DM yield per mm of used water, sorghum obtained similar values in both sowing methods, but above than maize at 11 and 39%, respectively.

Sorghum forage showed a total fiber content (FDN) lower than maize, which can mean a higher potential fodder consumption by livestock (^{Bolanos-Aguilar et al., 2013}). Improved forage quality by planting in wide beds of 1.6 m at six rows with water uptake in situ was possible in brown midrib sorghum where the total fiber content (FDN) was significantly lower (p< 0.05) to that of maize in both sowing methods; however, the FDA value did not show statistical difference, although on average it decreased by 2.6% compared to maize, both in traditional and 4-row edge wide beds. This indicates that sorghum produced forage with a higher proportion of digestible fiber than maize (^{Reta et al., 2010}).

Also, the increase in dry matter was due to a greater coverage of land, resulting from the better spatial arrangement that is obtained in the sorghum planting in wide beds at the edge of 1.6 m with six rows, respectively; which is an important management practice that increases the percentage of intercepted radiation, achieving a good canopy and a greater land coverage (^{Keating and Carberry, 1993}).

Conclussions

The climatic conditions of El Llano, Aguascalientes are suitable for the production of forage sorghum under rainfed conditions, since the thermal requirements of the crop are satisfied.

The cultivation of sorghum for forage planted in wide beds at the edge with wáter capture in situ with six rows is a viable option for rainwater shortages in the poor rainfed season in the region of El Llano, Aguascalientes, due to the higher efficiency in the use of this resource.

In the Llano region, Aguascalientes, although less heat is available for cultivation than in Mexico’s sorghum zones, the thermal oscillation is higher, allowing a good adaptation.

Brown midrib sorghum planted in wide beds at the edge with six rows and rainwater capture in situ showed higher dry matter yield and forage quality as well as better efficiency of rainwater use, than the maize planted in wide beds with four rows and single row.