Histological arrangements of plant tissue of different elephant grasses as influenced by their genotypes

Elephant grass genotypes display a variety of morphological differences, influencing the nutritive value of the forage. This study evaluated the histological arrangements of the leaves and stems of different elephant-grass genotypes, two tall-sized (Elephant B and IRI-381) and two dwarfs (Mott and Taiwan A-146 2.37), during a two-year trial. The grasses were harvested at 60-day intervals for two years. Biometric analyses of the stems and leaves were performed based on histological measurements. An in vitro dry matter digestibility (IVDMD) assay of the forage was performed. Among the elephant grass genotypes, the lignified cells and vascular bundles of the stems had higher variation than the leaves. Tall-sized genotypes displayed more lignified tissues in stems than the dwarfs. The transversal area occupied by vascular bundles and lignified cells were higher in Elephant B (44 911 μm2 and 35 895 μm2) (p < 0.05), compared to the dwarfs. Forage IVDMD was higher in leaves (699 g kg−1 of dry matter [DM]) than in the stems (678 g kg−1 of DM), considering all genotypes (p < 0.05). We did not observe any direct influence of genotype on forage digestibility, despite some differences in the histological arrangements.


Histological arrangements of plant tissue of different elephant grasses as influenced by their genotypes
Thaíse Virgínia Freire Ramos Peixoto 1,2  , Alexandre Carneiro Leão de Mello1,2 , Mércia Virginia Ferreira dos Santos 1,2  , Márcio Vieira da Cunha 1 , Rejane Magalhães de Mendonça Pimentel 1 , Luiz Henrique Gonçalves da Silva 1  , Djalma Euzébio Simões Filho 1 and Janerson José Coelho 1,3 * the lignified cells in forage plants affect their digestibility, by favouring or restricting access to cellular content by the rumen microorganisms (Sanchês et al. 2018).Therefore, understanding anatomical and histological patterns in different elephant grass genotypes can be a useful tool for identifying promising cultivars with good forage quality parameters.This could also help understand the factors limiting their consumption and digestibility.This study hypothesised that different elephant-grass genotypes display significant variations in their anatomical structures which can impact differences in their forage digestibility.Thus, the objective of this study was to evaluate the histological pattern of different elephant grass genotypes (dwarf and tall-sized) harvested at 60 days growth intervals and their influence on forage digestibility.

Material and methods
The experiment was carried out at the Estação Experimental de Cana de Açúcar do Carpina (EECAC), belonging to the Universidade Federal Rural de Pernambuco (UFRPE), located in Carpina city, Pernambuco state, Brazil.The average altitude is 180 m above sea level, with the latitude 7°51′03″ S and the longitude 35°15′17″ W. The climate of the region is As' or tropical dry (Köppen 2020), with the rainy season from May to September.The temperature and precipitation averages are 25 °C and 1 100 mm year -1 , respectively.The predominant soil is red-yellow ultisols (Santos et al. 2018), with slight topography.
The treatments were composed of four elephant-grass genotypes, two tall-sized  and two dwarfs (Mott and Taiwan A-146 2.37), harvested at 60 days growth intervals during a two-year trial (April 2015 to August 2017).The experimental design was a randomised complete block design, with four treatments (genotypes) and four replications (n = 16).The total area of the experimental plot was 25 m 2 (5 m × 5 m), with 9 m 2 (3 m × 3 m) designated as the sampling area.The first forage sampling for bromatological analysis was performed 60 days after establishing the plots in April 2015.For two consecutive years, forage samples were collected at 60day intervals, totalling 13 evaluations.The samples used for histological analysis were also harvested at 60 days of regrowth, and collected during the final evaluation cycle in August 2017.The samples were analysed in the Animal Nutrition Laboratory and Functional Phyto Morphology Laboratory located at UFRPE main campus.
The samples for histological analysis were rapidly kept in a FAA50 solution (formaldehyde, glacial acetic acid, 50% ethanol and 5:5: 90 v/v) at the harvest, aiming to preserve the forage quality (Johansen 1940).Fractions of the middle portion of the leaf blades from fully expanded leaves were sectioned before the analysis, the ligules were removed, and the length was measured from the apex to the base (Brito et al. 1999).In the stem, the samples were taken from the middle portion of the third internode, in the base-apex direction.The cross-sections of the leaf blade and stem were displayed and photographed using a Samsung CCD camera (with multi-magnification objective lenses) for further digital image analysis using Image Tool CMEIAS software (Wilcox et al. 2002) and an Opton microscope.
The biometric analyses of the leaves evaluated the following parameters: a) central rib -number of vascular bundles, number of lignified cells in the vascular bundles, cross-sectional area occupied by the lignified cells in the vascular bundle; b) mesophyll -number of vascular bundles in a 750 µm extension, the average number of lignified cells per vascular bundle of secondary veins, and the cross-sectional area occupied by lignified cells in the visualised vascular bundles.The measurements performed in the stems vascular bundles occurred in two distinct regions of the vascular cylinder: 1 -a more external area, with a greater occurrence of vascular bundles of different sizes and; 2 -an inner region with larger and more distant vascular bundles.In the digital images, the vascular cylinder regions were delimited with an approximate area of 775 × 103 m 2 , under a 4× objective.
Considering the locations described above, the biometric analyses in the stem focused on the total cross-sectional area occupied by the vascular bundles and the cross-sectional area occupied by lignified cells in the vascular bundle.In region 1, the percentage of the transverse area occupied by the lignified tissues closer to the epidermis/ outermost total area, and the percentage of the transversal area occupied by the parenchyma closest to the epidermis/ outermost total area were taken.In region 2, the percentage of the cross-sectional area occupied by the lignified tissues in the inner area/total innermost area and the percentage of cross-sectional area occupied by the parenchyma in the inner area/total innermost area were recorded.All biometric analyses were performed on digital images using the ImageJ/ Fiji program (Schindelin et al. 2012).
For in vitro dry matter digestibility (IVDMD), neutral detergent fibre (NDF), and crude protein analysis, the plants harvested were separated into leaf blade + sheath and stems, and dried at 55 °C to a constant weight.The samples were ground in a mill (Willey type) using a 2 mm sieve.The dry matter content of the samples was measured at 105 °C to a constant weight.NDF analysis was based on the methodology proposed by Van Soest (1991) and adapted by Senger et al. (2008), detailing the use of an autoclave at 110 °C for 40 minutes.The total nitrogen was estimated using the Kjeldahl method, the values were multiplied by 6.25 to obtain crude protein (CP) contents, according to Detmann et al. (2012).The equipment DAISY II Incubator (ANKOM ® Technology) was used for digestibility analysis.Artificial saliva was prepared according to McDougall (1948), consisting of 9.80 g l −1 NaHCO 3 ; 7.0 g l −1 Na 2 HPO 4 •7H 2 O; 0.57 g l −1 KCl; 0.47 g l −1 NaCl; 0.12 g l −1 MgSO 4 •7H 2 O; 0.05 g l −1 CaCl 2 •2H 2 O, making up a 1 000 ml volume with distilled water.The ruminal inoculum was composed of the solid and liquid fractions of the ruminal content collected from two adult sheep with a permanent cannula in their rumen.After a 48-hour incubation period, 40 ml of 1:1 solution of HCl (6 N) was added to each vessel to decrease pH to a value below 2; 8 g of pepsin was added and incubated for another 24 hours before estimating the IVDMD.The methodological procedure used to analyse the IVDMD was fully described in Holden (1999).
The histological parameters were evaluated by ANOVA, followed by the post-hoc Tukey's test (p < 0.05), using SPSS 23 IMB ® software.The in vitro dry matter digestibility, NDF, and CP data were analysed using the PROC MIXED procedure of SAS, considering blocks as random factors.The elephant grass genotypes and plant tissue (leaf and stem) were fixed factors.The averages were compared using the post-hoc Tukey's test (p < 0.05).Pearson's correlations were performed between histological variables and forage digestibility (p < 0.05).The following magnitudes of correlation were considered: if r < 0.20, non-existent correlation; 0.20 < r < 0.40, weak correlation; 0.40 < r < 0.60, moderate correlation; 0.60 < r < 0.80 strong correlation; if r > 0.80 very strong correlation.

Results
The biometric analysis of the leaves showed that only the area occupied by lignified cells in the main vein varied among the elephant grass cultivars (p < 0.05) (Table 1).The largest area was recorded for the genotype Elephant B (29 799 μm²) (p < 0.05), and the smallest area in Mott and Taiwan A-146 2.37 (16 867 and 17 696 μm², respectively).The number of vascular bundles ranged from 6 to 7, and the number of lignified cells in the mesophyll vascular bundles ranged from 33 to 53 (Table 1).The number of vascular bundles in the main vein varied between 11 and 22.The quantity of lignified cells in the central rib vascular bundles ranged from 120 to 182.The area occupied by lignified cells in the mesophyll varied from 5 764 to 11 532 μm².
The biometric analysis of the stems showed that three variables differed (p < 0.05) among the elephant grass genotypes (Table 2).The transversal area occupied by lignified cells in the vascular bundles was higher in Elephant B (35 895 μm²) (p < 0.05), in comparison to the dwarf genotypes Mott and Taiwan A-1.46 2.37 (16 630 and 12 671 μm², respectively).The transversal area occupied by vascular bundles was also higher in Elephant B (44 911 μm²) (p < 0.05), lower in Taiwan A-1.46 2.37 (18 011 μm²), while IRI-381 and Mott did record statistically similar values (p > 0.05).The percentage transversal area occupied by the parenchyma closest to the epidermis was higher in Taiwan A-1.46 2.37 (86%) (p < 0.05), lower in IRI-381 (77%), and Elephant B and Mott recorded similar values.In the elephant grass stems, the cell wall thickness ranged from 456 to 1 276 μm², and the number of lignified cells in the vascular bundles was between 194 and 276.The cross-sectional area occupied by parenchyma tissues varied between 89.3 to 96%.
All elephant grass leaves displayed homogeneous mesophyll, with five layers of parenchymatic cells, vascular bundles of different sizes distributed in parallel, and close to the fibre mesh close to the epidermis (Figure 1a,c,e,g).The leaf structure of the elephant-grass genotypes showed that tall-sized (Elephant B and IRI-381) and Mott had more visible chloroplast inside the bundle sheath than Taiwan A-146 2.37) (Figure 1).A greater density of vascular  bundles was observed in the outermost portion close to the epidermis in the stems of all genotypes (Figure 2).IRI-381 genotype had a higher presence of sclerenchyma sheath in the stem than the other genotypes (Figure 3d,e,f).
Comparing leaf and stem digestibility, there was no significant effect of the interaction of genotype and plant tissue (leaf and stem) (p = 0.29), and the main effect of genotype (0.07).There was a significant difference in the digestibility between the plant tissues isolated, with the leaves displaying a greater digestibility coefficient (699 g kg -1 of DM) in comparison to the stems (678 g kg -1 of DM) (p < 0.05).NDF contents were greater in tall-sized genotypes, in both leaves and stems (Table 3) (p < 0.05), only in the stems there were no significant differences between Taiwan A-146 2.37 and the tall genotypes.Taiwan A-146 2.37 had the highest crude protein content in the leaves (103 g kg -1 of DM), Mott had the lowest (92 g kg -1 of DM), while the tall genotypes recorded similar values.IRI-381 recorded the lowest crude protein content in the stems (48 g kg -1 of DM) among all genotypes evaluated.Correlation tests performed did not detect associations between the variations in plant tissues and the digestibility coefficient (p > 0.05).
Correlation results with weak correlations (r = 0.2 to 0.4), were not considered.It is worth mentioning that the range of digestibility recorded for the leaves and stems of the genotypes was low (67.3 to 70.4%).

Discussion
The histological comparison among different types of elephant grass genotypes, which included dwarfs and tall-sized, indicates that leaves had lower histological differences in the number of vascular bundles and lignified cells.This suggests that despite the variation in plant sizes and architecture, the leaf histological structure tends to be quite similar when these plants are harvested at 60-day intervals.This also relates to the observed narrow difference in leaf digestibility coefficients of the genotypes, with a low variation of 1.3%.These findings differed from the reports by Viana et al. (2018) who studied the same cultivars harvested at 32-day intervals.The authors found that the tall-sized genotypes had a higher proportion of lignified tissues in the On the other hand, the comparison of the histological structures of the stems revealed differences among the genotypes and sizes.The fact that the area occupied by vascular bundles and lignified cells were larger in tall-sized genotypes was possibly linked to their growth habit, especially in tall-sized genotypes.For instance, Elephant B had higher lignified structures in its stems than the dwarf genotypes (Viana et al. 2018;Souza et al. 2021).Elephant grass tall-sized genotypes are known for having a larger elongation of their internodes compared to dwarf genotypes.According to Yan et al. (2022), the extension of the cell wall is an essential factor for the plants to grow higher in elephant grass, and two primary components, cellulose and lignin, play a major role in this.Yan et al. (2021) studied the mechanisms of stem elongation in dwarf and tall-sized genotypes of elephant grass.The authors reported that the dwarf genotype Mott had down-regulation in expressing KS and GA20ox genes involved in the synthesis of gibberellin and lignin.Also, Mott showed up-regulation of the gene (GA2ox), which is involved in gibberellin inactivation.Despite the difference in the lignin content in the stems of different elephant grass genotypes in this study, these histological differences did not affect stem digestibility.
The leaves had higher digestibility than the stems.However, genotype did not influence stem digestibility, suggesting that forage digestibility coefficients in the elephant grass genotypes under 60-day harvesting frequency is closely related to the plant part analysed than the histological arrangement in the plant tissues.The elephant grass genotypes evaluated in this trial differ in terms of neutral detergent fibre (NDF) and crude protein (CP) (Peixoto 2018).In general, FDN values were higher in tall-sized elephant grass genotypes, ranging from 62 to 67% and 65 to 71% in the leaves and stem, respectively.Crude protein varied from 9-10% in leaves, and 4-6% in the stems, and higher values were recorded in the dwarf genotypes.
The current literature related to the nutritional value of different-sized elephant grass genotypes indicates that dwarf genotypes tend to have higher forage digestibility than tall-sized genotypes, usually a consequence of the higher leaf:stem ratio in dwarf genotypes (da Silva et al. 2021;Silva et al. 2021;Souza et al. 2021).Also, more lignified tissues in the stems of tall-sized genotypes are generally found as a consequence of their genetic predisposition for elongating the internodes more than the dwarfs (Viana et al. 2018;Souza et al. 2021;Yan et al. 2021).
The quantity, position, and degree of lignification of the vascular bundles in the tissues impact forage digestibility, as they act like a natural barrier impeding the rumen microorganisms to access the most digestible cells and tissues (Moore et al. 2020;Sanchês et al. 2021).In grass species, the vascular bundles (xylem and phloem) are associated with the sclerenchymatous tissues, which are very lignified, promoting plant support (Valente et al. 2016).The sclerenchyma fibres compose the vascular bundle sheaths in grasses (Matos et al. 2013), and the mature cells of this tissue are composed of dead cells with thick walls containing lignin, and minor physiological functions (Crang et al. 2018).The cells of the sclerenchyma and xylem are the most indigestible parts of the forage (Moore et al. 2020).Among the chemical components of the cell wall, lignin represents the main limitation to the digestion of the polysaccharides in plant cells (Valente et al. 2016;Moore et al. 2020).It is worth noting that the cells of the xylem and phloem, also undergo a secondary thickening of their walls (Moore et al. 2020).
Considering these facts, the closer the vascular bundles are to the outermost portion of the epidermis and closer to each other, the difficult it will be for the rumen microbes to break through these barriers to access digestible cells and tissues (e.g.mesophyll) (Brito et al. 2004;Sanchês et al. 2021).In the present study, the tall-sized elephant grass genotype Elephant B showed a larger area occupied by lignified cells in the leaf main vein, the transversal area occupied by the vascular bundles, and the area occupied by lignified cells in the stem vascular bundles than the dwarfs.This is an indication that the tissues of the tall-sized genotypes possibly have more structural barriers for the rumen microbes than the dwarfs.On the other hand, when there is a greater area of intercellular spaces in plant tissues, rumen microorganisms quickly access the most digestible cell walls, increasing the digestion rates of the forage (Paciullo et al. 2001).Less lignified tissues also facilitate the fragmentation of the forage by chewing (Paciullo 2002).In grass species, the vascular bundles are scattered throughout the whole fundamental tissue surrounded by a sclerenchymatous ring and covered by the epidermis (Matos et al. 2013;Heckwolf et al. 2015).During the early stages of plant growth, the xylem is the most lignified structure, but as the plant ages, there is a gradual lignification process in the sclerenchyma, a specialised support tissue (Akin 1989;Voxeur et al. 2015;Wang et al. 2019).Terry and Tilley (1964) reported that when the grass tillers are still very young, they can display digestibility coefficients close to or even greater than those found in their leaves, this is attributed to the greater content of soluble carbohydrates in young stems.However, in the very early stages of the development of the grasses, the tillers can mostly consist of leaves and pseudo-stems, and once the stems begin to elongate, the leaf:stem ratio can decline as the plant ages (Moore et al. 2020).Dry matter content also increases as the plant grows, a consequence of cell wall thickening and lignification, especially in the stems, which decreases forage digestibility (Tessema et al. 2010;Moore et al. 2020).Lista et al. (2020) reported negative correlations between the number of vascular bundles, percentage of xylem, phloem, and sclerenchyma with the IVDMD for different types of elephant grass; on the other hand, the percentage of parenchyma and mesophyll had a positive correlation with the digestibility.
The digestibility of the cell wall in forage plants is affected by the characteristics of the particles generated after the physical fragmentation by chewing, and also by the retention time in the rumen (Wilson 1994;Bruinenberg et al. 2002;Van Soest et al. 2019).Low digestion rates of some plant tissues are related to their physical aspects, especially cell wall thickness, but also their chemical characteristics, especially the degree of lignification (Valente et al. 2016;Moore et al. 2020).Nevertheless, forage digestibility will depend on a series of factors, including the interaction between rumen microbiota with the type of forage and feed particles (Pitta et al. 2010;Saro et al. 2012;Saro et al. 2014;Gruninger et al. 2019;Moore et al. 2020).In this context, the physical barriers promoted by the distribution of vascular bundles and lignified tissues, the cell arrangements, and chemical composition, are determinants but not the only source of variations affecting forage digestibility.

Conclusions
When harvested at 60 days of growth, different-sized elephant grass genotypes displayed more differences in their stem histology than the leaves.Tall-sized elephant grass genotypes tended to have more lignified tissues than dwarf genotypes, especially the stems.Genotype did not affect forage digestibility, despite the observed difference in histological arrangements.The findings of this study suggest that the digestibility of the four elephant grasses is less influenced by genotype than morphological fractions, with leaves having higher digestibility than stems.

Table 1 :
Histological biometrics parameters of the leaves of different elephant grass genotypes Means followed by different letters along the rows are significantly different in the post-hoc Tukey's test.SE = standard error (n = 16).Means followed by different letters along the rows are significantly different in the post-hoc Tukey's test.SE = standard error (n = 16).

Table 2 :
Histological stem biometrics parameters of elephant grass genotypes

Table 3 .
In vitro dry matter digestibility (IVDMD), neutral detergent fibre (NDF), and crude protein (CP) (g kg -1 of dry matter) of the leaf and stem, in different elephant-grass genotypes