Forage sorghum is a viable alternative for producing silage on dairy farms located in regions with arid and semi-arid environments. This crop has been shown to grow well under conditions of limited water ¹ and high temperatures²; in addition to having moderate tolerance to soil salinity³. Under these conditions, sorghum has the advantage of being able to produce larger amounts of dry matter (DM) compared to forage maize^4,5. However, the lignin content in the forage of conventional sorghum varieties (up to 9.1 % of DM) is associated with silages that have low NDF digestibility (NDFD); therefore, when used in the diets of dairy cows, it limits the consumption of DM and milk production^6,7.

A practical option to reduce lignin concentration and improve the digestibility of the silage fiber of different forages is to increase the cutting height. In this regard, it has been reported that increasing the cutting height from 15 to 45 cm in black sorghum (Sorghum almum) reduces lignin contents from 7.7 to 6.4 %⁸. In forage maize, it was found that lignin decreased from 3.0 to 2.6 %, while NDFD increased 2.3 % when cutting height at harvest increased from 12 to 45 cm⁹. Similarly, the increase in cutting height in forage maize from 15 to 45 cm increased NDFD by 5.0 %, which was attributed to decreasing the proportion of basal stems containing the most lignified part of the plant¹⁰.

When the cutting height of the forage increases, the nutrient composition of silage improves; however, DM yield losses can become very significant if very high cutting heights are used. The above causes that this harvesting strategy is not accepted by producers. Yield losses in DM of sorghum forage can be between 10 and 20 % when increasing the cutting height above 20 cm^11,12 and from 3 to 16 % when increasing the cutting height to more than 40 cm in forage maize^10,13. Therefore, it is necessary to identify the best cutting height in forage sorghum, which allows having an optimal balance between the nutritional composition of silage and the DM yield per hectare at harvest. The objective of the present work was to identify the optimal cutting height in forage sorghum to improve the nutritional quality of silage without affecting the DM yield.

The experiment was established during the spring 2018 production cycle in the Ejido Venecia, located in the Municipality of Gómez Palacio, Durango. The experimental site is located at 25º46'56" N and 103º21'02" W at an altitude of 1,100 m above sea level. The soil has a clay texture, with a bulk density of 1.07 g cm^-3, an organic matter content of 1.5 % and a pH of 8.3.

Precipitation and air temperatures during crop development are shown in Figure 1. The accumulated precipitation during the cycle was 22.4 mm. Maximum temperatures ranged from 28.7 to 43.1 °C and minimum temperatures from 13.1 to 27 °C. The highest temperatures that exceeded the optimal growth range in sorghum (6 - 37.7 °C)³ occurred between 41 and 52 d after sowing (DAS).

Figure 1 Maximum, minimum and average temperature and precipitation recorded during the experimental period (April 18 to August 10, 2018) 

The effect of six cutting heights (10, 20, 30, 40, 50 and 60 cm from the soil surface) on forage yield, nutrient composition and nutrient yield of forage sorghum silage was evaluated. The preparation of the land consisted of a fallow, double harrowing and leveling with scraper. The sowing was carried out in moist soil on April 18, 2018 with a Gaspardo precision seeder (model SPLC-4F) using a sowing density of 12 kg ha^-1 of seed of the Silo Miel variety (Agricenter Zevilla, Torreón, Coahuila). Four furrows per treatment of 0.8 m in length at a distance of 0.75 m were sown using a completely randomized block design with four repetitions. There was an average density of 195,000 plants ha^-1. The fertilization was carried out with 134 kg ha^-1 of N and 43 kg ha^-1 of P₂O₅. The total phosphorus was applied at sowing, while the N was fractionated, applying 40 % of the total dose at sowing and 60 % before the first supplemental irrigation; 48 DAS. At sowing, 8 kg ha^-1 of K and 10 kg ha^-1 of S were also applied and, before the first supplemental irrigation, 18 kg ha^-1 of Ca and 12 kg ha^-1 of Mg were applied. Yara Mila Star® and Yara Bela Nitromag® were used as fertilizer sources (Yara, Guadalajara, Jalisco). Four irrigations were applied including a pre-sowing irrigation and three supplemental irrigations at 48, 65 and 85 DAS. A furrow irrigation system was used. At 25 and 46 DAS, applications of chlorpyrifos ethyl (Lorsban 480 EM®, BASF Inc., Germany) were made at the rate of 0.75 L ha^-1 for the control of fall armyworm (Spodoptera frugiperda). Subsequently, at 50 and 83 DAS, applications were made for the control of yellow sugarcane aphids (Melanaphis sacchari) using Imidacloprid + Betacyfluthrin (Muralla Max®, Bayer, Mexico) and Sulfoxaflor (Toretto Isoclast® Active, Corteva Agrosciences, Guadalajara) at a rate of 0.25 ml ha^-1 and 100 ml ha^-1, respectively. Weed control was performed manually.

The harvest of the crop was carried out on August 10, 2018, at 105 DAS when the grain reached the milky-dough stage, accumulating 2,118 heat-hours. The two central furrows were used as a useful plot, removing 1 m from each end to exclude the edge effect. In total, six meters in length for each of the treatments (9.12 m²) were harvested. Each useful plot was harvested considering the different cutting heights of each treatment (10, 20, 30, 40, 50 and 60 cm) based on the soil surface. Fresh forage of each useful plot was weighed to estimate the green forage yield. From the total plants cut per plot, 15 plants were randomly selected and ground to a theoretical particle size of 2 cm using a mill Model JF5. From the fresh ground forage of each plot, three random samples of 500 g each were taken and dried at 60 °C until constant weight in a forced air stove to determine the DM content. The DM yield was determined by multiplying the green-based forage yield per hectare by the DM content of the forage before silage.

To make the mini silos, the first three repetitions of the chopped fresh forage of each treatment were used. Glass jars with an airtight lid of 1 L capacity were used, where the chopped fresh forage was compacted to a density of 261 kg of DM m^-314 in each mini silo considering the content of DM at the harvest of each treatment. The estimation of the DM content to determine the density in the mini silos was made with the microwave oven and an average DM percentage of 29.67 ± 0.42 % was estimated. Compaction of the forage in each mini silo was carried out manually with a macho meat tenderizer (Metaltex 779-012). All mini silos were stored in the laboratory at ambient temperature for 90 d.

When opening the mini silos, the first 5 cm were discarded. Subsequently, in each mini silo a sample of 20 g of fresh silage was taken to which 200 ml of distilled-deionized water was added and it was mixed for 30 sec in a high-speed blender. The diluted sample was filtered through three layers of cheese gauze and the pH was measured in the liquid with a portable potentiometer (OHAUS Model ST2100, Parsippany, NJ, USA)¹⁵. From the remnant of the fresh silage, 400 g was taken and sent to a private commercial laboratory (GAQSA, Querétaro, Mexico) to be analyzed using near-infrared spectroscopy (NIR; Mod. 951, Foss Electric, Hillerod, Denmark). Each fresh silage sample was homogenized and dried in an airflow furnace at 66.7 °C until constant weight. Subsequently, the samples were crushed in a blender, and consecutively, in a mill using a 1 mm mesh. In these samples, the contents of crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF), lignin, in vitro digestibility of NDF at 30 h (NDFD-30h), TND, NFC, NEL and ashes were analyzed. The determination of nutritional values was based on prediction equations and databases generated by the “Cumberland Valley Analytical Services (CVAS)”. The calibration of the equipment considers the following procedure: selection of the sample at random, taking of the spectra of the sample, selection of the spectra representing the samples, laboratory analysis of the selected sample using the reference methodology, comparison of the results of the reference methodology with their respective spectra, internal validation with the samples and carrying out the analysis by the reference methodology and validation of the calibration equation with the reference results of the internal validation samples. Once the calibration equation with the reference results of the internal validation was satisfactory, it was used for the analysis of all samples.

The yields of NDF, NDFD-30h, TDN and NEL were estimated considering the content of these nutrients in the silage and the dry forage yield per hectare of each treatment.

The analysis of the information was carried out with the statistical program SAS version 9.3 (SAS Institute Inc., Cary, NC. USA). The results were analyzed by ANOVA using a randomized block design, with six treatments, of four and three repetitions for the variables of forage yield and silage quality, respectively. When significant differences were found (P˂0.05), Fisher's protected least significant difference test was applied to compare means between treatments at the same level of significance.

Plant height after applying the treatments and forage yields are shown in Table 1. As expected, the plant height at harvest was higher when a cutting height of 10 cm was used and lower when the forage was harvested at 60 cm. Yield results indicated that fresh and DM productions can decrease from a cutting height of 40 and 50 cm, respectively, from the soil surface. These reductions can be considerable up to 12 % when the cutting height reaches 60 cm compared to cutting heights between 10 and 40 cm. In this regard, other works have reported that the DM yield reduces as the cutting height increases, and that the losses in yield can range from 10 to 20 % when the cutting height is greater than 20 cm in forage sorghum^11,12.

The nutritional value of sorghum silage altered as the cutting height increased (Table 2). The NFD content reduced from 74.1 % to 66.9 % when increasing the cutting height from 10 to 20 cm, respectively, but no significant reduction was observed from 20 cm to the cutting height of 60 cm. The lignin content reduced from 8.1 % to 6.4 % when the cutting height increased from 10 to 30 cm, but there were no significant changes from 30 cm to the cutting height of 60 cm.

As a result of the observed changes in NDF and lignin contents, NDFD and the TDN concentration of the silage may increase after the forage was harvested at cutting heights of 40 and 20 cm, respectively. Similarly, the NFC content of sorghum silage increased as the cutting height increased, where the treatments of cutting height of 50 and 60 cm were the ones with the highest concentration of carbohydrates. This is probably associated with the lower lignification of the fiber and its greater digestibility, which in turn influenced the increase in the concentration of NEL of the silage from the cutting height of 40 cm^6,7.

The reduction in lignin content observed in the present study had a positive impact by improving fiber digestibility and energy availability in silage, which has been a strategy to improve the digestibility of fibrous forages¹⁶. In a study where the cutting height was increased from 15 to 45 cm in the forage of Creole maize, black sorghum and King grass, the concentrations of lignin and NDF in maize and grass did not change; however, the highest cutting height reduced the concentration of NDF by 4.2 % and lignin by 1.3 % in sorghum forage⁸. In another similar study, in forage maize, increasing cutting height from 12 to 45 cm at harvest reduced lignin content by 0.5 %, which contributed to the increase in fiber digestibility by 2.4 % in maize silage⁹. It has been reported that for every percentage unit that NDFD increases in forage, DM consumption and milk production in cows increases by 0.17 and 0.25 kg d^-1, respectively¹⁷.

Regarding nutrient production, it was found that the NDF yield decreased as the cutting height of forage at harvest increased (Table 2). This is due to the decrease in the concentration of NDF of silage and the increase in yield of DM of forage as the cutting height of the forage increases. However, the cutting height did not affect NDFD, TDN NEL NLE yields.

The pH of sorghum silage in response to the cutting height of the forage at harvest is shown in Figure 2. Cutting sorghum forage at a height of 30 cm from the base of the soil could favor better fermentation of silage, as a lower pH was observed (3.9). In contrast, silage of sorghum forage cut at a height of 20 cm registers the highest value (4.8). In relation to the other cutting heights, 10, 40, 50 and 60 cm were not statistically different, and cutting at 10 cm could lead to a pH in silage similar to harvesting forage at 30 cm. The final pH of silage can be affected by many factors, but it is mostly related to the concentration of carbohydrates in the forage¹³. In the present study, the concentration of NFC in silage begins to increase after the forage was cut at 30 cm (Table 2), which coincides with the reduced pH of silage in this treatment. However, it is important to analyze, in future studies, other organic compounds and final products of fermentation that help confirm this effect. In forage maize, increasing the cutting height of the forage from 12 to 45 cm did not change the pH of the silage but increased the concentration of NFC such as starch⁹.

Means with different letters are significantly different (P˂0.05).

Figure 2 pH of sorghum silage in response to cutting height in sorghum forage at harvest 

In conclusion, by cutting the sorghum forage between 20 and 40 cm from the soil surface, a silage with lower lignin content, greater fiber digestibility and good energy content was obtained, without compromising the yield of DM per hectare. In addition, harvesting the forage 30 cm from the base of the soil promotes a good fermentation of the silage. Therefore, increasing the cutting height of sorghum forage up to 40 cm above the soil basis improves the nutritional quality of silage without significantly reducing the DM yield.