Bromatological and histological features of native African grasses under grazing in Brazilian semi-arid rangelands

This study evaluated the bromatological and histological features of native African grasses under grazing in Brazilian semi-arid rangelands. An experimental grid design was used in a randomised sampling method for four replicate samples of three African grasses. The grasses evaluated were: Cenchrus ciliaris L., Digitaria pentzii Stent. and Megathyrsus maximus (Jacq.). The bromatological and histological traits analysed included dry matter (DM), crude protein (CP), neutral detergent fibre corrected to ashes and protein (NDFap), lignin, mesophyll, vascular bundle, phloem, xylem, vascular sheath, sclerenchyma, bulliform cells, and the adaxial and abaxial epidermis. The type of grass and grazing cycle correlated with DM, NDF and lignin (p < 0.05), whereas CP was not affected by these factors (p > 0.05). Dry matter ranged between 128 and 304 g kg−1; CP ranged between 90 and 167 g kg−1; NDFap between 542 and 707 g kg−1, and lignin between 10 and 40 g kg−1. The proportion of xylem (8.4%), vascular sheath (30.5%) and total vascular bundles (38.1%) were greater in M. maximus (p < 0.05). The percentage of sclerenchyma did not differ between species (p > 0.05). Digitaria pentzii and C. ciliaris had a greater proportion of mesophyll than M. maximus (p < 0.05). Of the three grasses analysed. Digitaria pentzii exhibited the highest forage quality, because it had the lowest concentration of fibre and lignin, with lower proportions of lignified tissues.

Three native African grasses were evaluated: C. ciliaris, D. pentzii and M. maximus (Supplemental Figure 1). Four replicate samples were randomly selected from an experimental grid design (n = 4). Each plot had a total area of 25 m 2 and a sampling area of 9 m 2 . The exotic African grasses were established in June 2009. However, only in September of 2010 the trial started by cutting all of them to the same stubble height of 10 cm to obtain uniform stands in September 2010. At the time in 2009, they were only planted. All of these were African grasses, therefore exotic. In April 2011, the plots were fertilised with urea at 60 kg ha −1 N prior to onset of grazing. The plots were periodically grazed to a post-grazing residual height of 20 cm by dairy cows at the stocking rate of one animal unit per plot, taking only a few hours to reach the desirable residual height. The grazing started in May 2011 and continued to September 2011, with a 35-day interval between each grazing cycle.
Pluck samples were collected for bromatological analysis prior to the grazing of each pasture (Edvan et al. 2016). Samples were dried in an air-forced oven at 55 °C for 72 hours to determine dry matter (DM) content. Later these samples were ground in a Wiley mill (MO6666, John Doe Co., Dog City, USA) using a 1-mm sieve.
For the histological analysis, fully expanded leaves were collected from three representative tillers of each plot. Immediately after tiller sampling, the third fully expanded leaf was taken and a median section (length = 3 cm) was fixed for tissue preservation using 50% formaldehyde alcohol acetic acid solution. The leaf blades were subjected to a progressive cyclic series of alcohol dehydration, followed by inclusion in Paraplast. Median portions (10 μm) of the leaves were transversely sectioned using a rotating microtome. The material was thereafter deparaffinised, and a quadruple tissue staining was performed using FASGA staining, as described by Tolivia and Tolivia (1987). Lignins, suberins and/or cutins were positively identified through the formation of a red stain following the reaction of safranin O with these phenolic compounds. Permanent histological slides were assembled for microscope analysis and long-term storage (Hagquist 1974). The measurements of leaf tissues were made using an optical microscope, with a coupled camera (Leica ® DM500, Leica Microsystems, Sao Paulo, Brazil). The images were analysed using Sigma ScanPro 5 ® software. The height and width measurements were performed in the region of the mesophyll, including, a whole vascular bundle. The areas of phloem, xylem, vascular sheath, sclerenchyma, total vascular bundle, buliform cells, mesophyll, and adaxial epidermis and abaxial epidermis were measured. Additionally, the numbers of total vascular bundles were recorded. Finally, the proportions of the tissues were calculated, counting ten observations per block of each species, totalling 40 images per species.
The analyses of DM content (930.15 g kg −1 ) and crude protein (CP) (984.13 g kg −1 ) were performed according to the Association of Official Agricultural Chemists (AOAC) (2005). The analysis of neutral detergent fibre (NDF) corrected to ashes and protein (NDFap) was performed according to the methodology of Van Soest et al. (1991) with modifications proposed by Senger et al. (2008), detailing the use of an autoclave at 110 °C for 40 minutes. For the determination of lignin (LIG), the acid detergent fibre (ADF) residues within of each sample bag (TNT, a non-woven fabric) were washed with a 72% sulfuric acid solution for 3 hours, with homogenisation occurring every 30 minutes to solubilise the cellulose and obtain the acid-digested lignin (Van Soest et al. 1991). The bags were then washed and weighed.
The data were tested for normality and homoscedasticity before the analysis of variance (ANOVA). The analysis of the bromatological data was performed using the procedure PROC MIXED of the statistical package SAS ® (Statistical Analysis System; SAS 2011). Bromatological variables were tested using a matrix of covariance symmetry (CS), and the comparison was performed by the LSMEANS adjusted to Tukey test 5% (p < 0.05). Histological data were tested using PROC ANOVA, and means were compared by the Tukey test 5% (p < 0.05).
The simulated grazing during the grazing cycles affected the bromatological variables, DM, NDFap and lignin, of the grass species (p < 0.05; Table 1). The DM content in the grasses ranged between 128 and 304 g kg −1 , where the lowest value was found in D. pentzii during the grazing cycle of June 2011 and the greatest DM concentration was found in C. ciliaris in May 2011. All the three grasses exhibited the greatest NDFap values in the first grazing cycle (May 2011), which ranged from 686 and 707 g kg −1 DM. Megathyrsus maximus and C. ciliaris had a higher NDFap concentration than D. pentzii throughout all the grazing cycles. Digitaria pentzii, however, showed the lowest NDFap concentration in June 2011 (542 g kg −1 DM). Lignin concentrations ranged between 10 and 40 g kg −1 DM across all grazing cycles, with M. maximus and C. ciliaris displaying greater concentrations than D. pentzii (p < 0.05). Crude protein concentrations ranged between 90 and 167 g kg −1 DM and did not differ significantly among the grasses (p > 0.05; Table 1).
Different exotic forage grasses growing in the Brazilian semi-arid rangeland show a varied pattern for most of their bromatological and histological traits evaluated, especially because these species had different morphophysiological characteristics and growth habits. Variations in plant quality among different forage plants arise primarily from the fact each forage species has unique morphophysiological features, which provide them with different adaption mechanisms to respond to the environmental variables (e.g., water and nutrient availability, radiation and light density, grazing intensity), which affect their growth and quality of the forage produced (Nelson and Moser 1994). Majority of the bromatological variables in this research were influenced by the grazing cycle, possibly as a response to the variations of the availability of water and nutrients between grazing events, and also the residual effects of the consecutive  27 Standard error = 0.14 Note: Within each variable, means followed by the same uppercase letter in the column and lowercase in the lines, are not significantly different by the LSMEANS test (p < 0.05). * Neutral detergent fibre corrected to ashes and protein  Adesogan et al. (2009), forage quality usually tends to decline after consecutive harvestings, because there is an accumulation of stems in the plants, as well as deposition of lignin in the leaves and stems that remain after grazing. Interactions between the grazing cycle and species were noted for all bromatological variables, except for CP.
In our present study, C. ciliaris and M. maximus showed increases in the lignin concentrations subsequent to the consecutive grazing cycles. Dry periods are also known to increase DM concentrations in forage plants (Coêlho et al. 2018). Studying these same grass species under two management conditions (deferment and grazing), Coêlho et al. (2018) reported concentrations of DM ranging between 318 and 370 g kg −1 in C. ciliaris, 221 and 299 g kg −1 in D. pentzii and 233 and 340 g in M. maximus; a greater range than found in our present study.
The average values of CP found in the forages (>90 g kg −1 DM) were considered sufficient to meet the minimum requirements for ruminants (70 g kg −1 DM). The phenotypical differences among the grass species studied, combined with grazing and environmental effects, did not lead to significant differences in crude protein content in this research. Coêlho et al. (2018), however, reported differences in CP content among C. ciliaris, D. pentzii, and M. maximus in grazed conditions, although they did not observe the same difference when these grasses were under deferment. Coêlho et al. (2018) also observed that NDF content among C. ciliaris, D. pentzii, and M. maximus in grazed conditions differed. Digitaria pentzii typically has a lower NDF content, compared with the other grasses, which is possibly associated with the phenotypical features intrinsic of this plant, especially the greater leaf: stem ratio and lower proportions of lignified tissues. The fact that all the grasses showed greater content of NDF in the first grazing cycles, where the precipitation was greater (Figure 1), might be associated with the faster growth during the wetter months, where most of the non-structural carbohydrates produced were used for fibre synthesis (Purdy et al. 2015). The proportion of NDF is related to the maximum consumption of dry matter, and plants with greater concentrations (>550 g kg −1 DM) can cause limitations in food intake and, consequently, energy consumption (Arelovich et al. 2008).
The tendency for greater concentrations of lignin in M. maximus and C. ciliaris than in D. pentzii can also be associated with the phenotypical features of these plants, because there was no defined pattern for the effect of the grazing cycle on this component. Megathyrsus maximus and C. ciliaris displayed a tendency to have greater proportions of lignified vascular tissues (xylem and total vascular bundles) than D. pentzii, which probably contributed to a greater concentration of lignin. Xylem, vascular bundles and sclerenchyma are associated with the deposition of lignin in forage tissues owing to the formation of secondary cell walls, which makes them difficult to digest (Giordano et al. 2014;Valente et al. 2016). The increased lignin concentration in some of the grass cell walls during the maturation phase decreases cell numbers, consequently increasing the difficulty of ruminal microbiota to digest these plants (Jung et al. 2012). In grasses, lignin is not a nutritional component and reduces the nutritional availability of the plant fibre, consequently interfering with the digestion of the cell wall polysaccharides by acting as a physical barrier to the microbial enzymes (Liu et al. 2018).
In terms of digestibility of the plant tissues, mesophyll and phloem are, however, quickly digested, whereas the epidermis and vascular sheaths have slower degradation (Wilson 1993;Valente et al. 2016).Digitaria pentzii and C. ciliaris had a greater proportion of mesophyll than M. maximus. According to Valente et al. (2016), the tissues of the mesophyll are not lignified, because they are primarily used for photosynthesis and therefore have elevated concentrations of metabolites and are easily digested.
A greater proportion of only the adaxial epidermis in D. pentzii in comparison to the other species was noted. Few other differences in the epidermis were recorded. The cells of the sclerenchyma are generally very lignified with thick walls (Tsuzukibashi et al. 2016), and because there were no significant differences in the proportion of sclerenchyma among the grasses, most of the differences that contributed to variations in the nutritional value among the grasses came from the proportions of total vascular bundles, vascular sheaths and xylem, especially comparing M. maximus and D. pentzii.
The three native African grasses tested could provide a relatively good quality of forage for ruminants in the Brazilian semi-arid rangelands. Most of the bromatological Means followed by the same letter in the lines are not significantly different by the Tukey test (p < 0.05). CV = coefficient of variation. Samples were collected at the end of the grazing season of 2011 Table 2: Tissue proportions (%) and the number of vascular bundles in leaf blades of exotic African grasses under grazing in the Brazilian semi-arid region and histological differences were attributed to the inherent characteristics of each forage species, even though there was an influence of the grazing events in modifying some of the bromatological features. Digitaria pentzii had the ideal bromatological composition, with a greater proportion of mesophilic cells and fewer indigestible tissues.