THE CHEMICAL COMPOSITION AND NUTRITIVE VALUE OF Brachiaria GRASS CULTIVARS AT KATUMANI DRYLAND RESEARCH STATION IN SOUTH EASTERN KENYA

The study was conducted at Kenya Agricultural and Livestock Research Organisation, Katumani, Machakos to evaluate the herbage quality of Brachiaria grass cultivars in semi arid regions of Eastern Kenya. Brachiaria cultivars namely B. decumbens cv. Basilisk, Brachiaria hybrid Mulato II, four Brachiaria brizantha cultivars Marandu, Xaraes, Piata, MG4 and Brachiaria humidicola cv Llanero were assessed with reference to their chemical and nutritive composition at 22, 24 and 28 weeks post seedling emergence. Rhodes grass (Chloris gayana cv KATR3) and Napier grass (P. pupureum cv. Kakamega I) were included as controls. There were significant differences (p<0.05) among the cultivars for crude protein, crude fibre (NDF, ADF and ADL), Ash, Calcium, Phosphorus, Dry Matter Digestibilty and Metabolisable energy during the the week 22 and 24 post seedling emergence harvest. At the week 28 harvest interval, however, ADL values for all the cultivars were similar. Only Marandu was significantly different (p<0.05) for values of metabolizable energy from the rest. The mean CP content of the grasses decreased from 11.1% at 6.3% from week 22 to week 28 harvest interval. MG4, Mulato II and Xaraes were the only grasses able to meet minimum CP (7.0%) requirement for rumen microbial function throughout the harvest period. Ash and phosphorus values decreased whereas Calcium content increased for all the cultivars during the harvest intervals, with Mulato II recording the highest ash (15%) content during this period. Chloris gayana KATR3 recorded highest average NDF (72.9%), ADF (48.1%) and ADL (6.1%) content during this period. Xaraes, Marandu and Mulato II were the only cultivars able to achieve the highest metabolizable energy of 7.0 MJ/kg DM recorded for the grass cultivars throughout the harvest intervals. These results indicate that Brachiaria grasses can be a good source of forage for livestock and a boost to the forage resource base in the semi arid regions of Kenya. Further research is needed to quantify their productivity in both dry and wet periods and to assess the effect of feeding on animal production performance. Indexing terms/

The nutritional value of tropical grasses has been shown to be generally low, falling rapidly with plant maturity [21]. Semi arid and arid rangelands are usually dominated by a particular forage type which provides the majority of organic matter consumed by grazing animals. The forage is relatively high in quality during early vegetative growth but quickly declines in quality as it matures. Their content of structural carbohydrates is quite high, they enter dormancy during unfavorable periods and reinitiate growth during favorable periods. Indigenous species that have evolved under the harsh climatic conditions of the ASALs in southern and central-northern rangelands of Kenya include grasses like C. ciliaris, E. superba, E. macrostachyus, and C. roxburghiana [22]. These grasses' nutritional and yield status decline with changing climatic conditions in the year. The full potential of the ASALs for livestock production can be exploited by expanding the forage resource base through introduction of climate smart forage species which will be able to boost nutrient quality and quantity hence supplying the nutritive requirements of livestock. Studies on climate smart Brachiaria grass species which originated from Africa and have been developed elsewhere could be the key to improvement of livestock production and could also serve as boost to the forage resource base in the ASAL regions. This study is aimed at evaluating the chemical composition and nutritive values of 7 Brachiaria cultivars namely B. brizantha cultivars Piata, Xaraes, MG4 and Marandu; B. decumbens cv. Basilisk; Brachiaria hybrid cv. Mulato II and B. humidicola cv. Llanero with Rhodes grass (C. gayana KATR3) and Napier grass(P. pupureum cv. Kakamega I) to serve as controls. J a n u a r y 08, 2 0 1 6

Site
The study was conducted at the Kenya Agricultural and Livestock Research Organisation, Katumani, Machakos, Kenya (10 58´S, 37 0 28´E). Elevation is 1600m above sea level and the mean temperature is19.6 0 C [23]. The mean annual rainfall is 717mm, with a bimodal pattern, the long rains (LR) occurring from March-May and the short rains (SR) from October-December with two dry seasons (June-September; January-February). The soils are Luvisols, low in nitrogen and phosphorus with a PH of 6.5 [24].

Experiment: Analysis of Chemical Composition and Nutritive Value of Brachiaria
Grass Cultivars harvested at 22, 24 and 28 weeks post seedling emergence.
Grass samples from a field sown in November, 2013 were harvested at week 22, 24 and 28 post seedling emergence from an area of 2m 2 in a split plot arrangement as in shown figure 1. The field comprised of 9 Brachiaria grass cultivars, namely; B. brizantha cultivars Marandu, Xaraes, Piata and MG4; B. decumbens cv. Basilisk; B. humidicola cv. Llanero and Brachiaria hybrid cv. Mulato II with Napier grass (P. purpureum cv Kakamega I) and C. gayana cv KATR3 serving as controls planted in a randomized complete block design layout with 4 replications. The individual plot sizes measured 5m x 4m with a 1m path between plots and 1m path between the replications. Prior to this, the grasses underwent standardization cuts at week 16 (onset of the March-May Long rainy season of the year 2014) post seedling emergence, marking the end of their establishment phase after which the plots were top dressed with CAN at rates of 100kg N/Ha. Napier was harvested only at week 24 post seedling emergence in an area of 4m 2 . A fresh weight of all the harvested material was recorded after which sub samples of these were weighed and recorded. The sub samples were dried in an oven for 72 hours at temperatures of 65 0 C after which the dry sample weights were recorded. These oven dried samples included the leaves and stems harvested at 5cm stubble height. The dried samples were then grinded with a hammer mill and sieved using 1mm sieves and stored in labeled plastic bags in preparation for the proximate analyses method [25] and the [26] feed analysis for feed quality assessment and chemical analysis.

Rep 1. split plot harvest arrangement
Rep2. split plot harvest arrangement

wks
Rep4. split plot harvest arrangement

Forage Quality Analysis.
Chemical analyses were carried out at the KALRO Muguga animal nutrition laboratory. The chemical analyses were performed to determine the, crude protein (CP), In vitro dry matter digestibility (IVDMD), Ash, Crude fibre (neutral detergent fiber (NDF), acid detergent fiber (ADF) and acid detergent Lignin (ADL)). Crude protein was analyzed using Kjeldahl nitrogen determination [27]. Fibre analysis was done using the Ankom fibre method which is a modification of the [28] of forage analysis. Each sample of the extractions was done in the order of NDF (%), ADF (%) and then ADL (%). For determination of in vitro dry matter digestibility (IVDMD), the two stage technique for in vitro digestion of forage crops [29] was used. The ruminal fluid was obtained from one rumen cannulated steer with an average weight of 550 kg kept on Rhodes grass pasture and hay. Ash was determined by weighing the resulting inorganic residue [25]. Minerals were analyzed on a Thermo ICAP 6300 Inductively Coupled Plasma (ICP) Radial Spectrometer. Metabolizable energy was estimated on the basis of the equation described by [30].

Statistical analyses.
Data on chemical composition, nutritive value and dry matter yields of forage samples were subjected to ANOVA based on the model designed for a randomized complete block design (RCBD) according to [31].To compare significant differences in response variables, ANOVA analysis was done using SAS package [32]. Duncan's Multiple Range Test was carried out for subsequent comparison of means as described by [33].

Climatic Data
Rainfall data (

Feed Quality and Chemical analyses
Mean ash content for the cultivars was significantly different at week 22 as shown in    Column means with similar superscripts are not significantly (p<0.05) different J a n u a r y 08, 2 0 1 6    MJ/KgDM. J a n u a r y 08, 2 0 1 6

DISCUSSION
With advancing stage of development during the harvest intervals, consistent decline in ash, phosphorus, Crude protein, IVDMD and metabolisable energy was observed. On the other hand Ca, NDF, ADF and ADL increased.
In this study, Chloris gayana cv KATR3 recorded highest calcium content during all the harvest intervals with Marandu and Mulato II recording highest phosphorus content. A high calcium content with increasing age can be explained on the basis of the increased amount of cellular material which is composed principally of this element [34]. [35] However reported that the late-season increases in calcium and ash may be attributed to dust accumulations. A late stage pregnant cow requires 11% of CP, 0.37% of Ca and 0.26% of P daily [36]. A dry cow requires 0.25% Ca and 0.16% P, while a lactating cow needs 0.31% Ca and 0.21% P. Among the grasses only Chloris gayana cv KATR3 is able to meet the calcium content of 0.37% for late stage pregnant cows. None of the grasses attained the Phosphorus content required by late pregnancy, dry and lactating cow.
For all the grasses the levels of CP in the harvested forage exceeded the minimum of 7.0% suggested as necessary for optimum rumen function by [37] at the week 22 harvest but dropped during the week 24 and 28 harvests. This trend in CP content has been reported in other studies by [38]; [39] and [40] and is mainly attributable to dilution of the CP contents of the forage crops by the rapid accumulation of cell wall carbohydrates at the latter stages of growth [37]. Comparative studies were developed by [41] who evaluated the nutritional values of Marandu palisadegrass, Xaraes palisadegrass and Piata palisadegrass and showed that regardless of the experimental year, the CP were higher during the rainy season. [42] reported that other factors like maturity of the grass can be the source of declining CP.
The Brachiaria brizantha cultivars had high (11.1-11.9%) and similar CP content with Mulato II (12.8%) at the week 22 harvest which can be partly attributed to the Nitrogen application of fertilizer at rates of 5.8 Kg N /Ha after standardization cuts were done on the plots. Studies carried out by [43] report that the effects of 3 cutting intervals (20, 40, 60days) and four rates of N application (0, 100,200 or 400kg N/ha) on the yield and CP content of B. brizantha at Morogoro Tanzania. Crude protein content of the herbage declined with longer cutting intervals, but increased from 6.9% to 12.9% when N was increased from 0 to 400kg/ha. Crude protein yield increased with increasing cutting interval up to the 40-day interval. It was concluded that at Morogoro, Tanzania B. brizantha should be harvested at about 6-week intervals to strike a balance between herbage yield and quality. Values for Napier (7.6%) which was cut once at week 24 were above 7.0% as recommended by [37] for rumen microbial function. In all the harvest intervals Mulato II recorded high CP levels that ranged between 7-12.8% which is almost within the range reported by [44] of 10-14% crude protein in Thailand on poor soils and 12-17 % crude protein on better soils in Florida, USA. Values for CP for Basilisk ranged between 9.8% (week 22 harvest) and 4.9% (week 28 harvest).These were lower than those found by [45] in the tropical region of Indonesia for Brachiaria decumbens collected in the natural grassland of Sumatra during the wet and dry seasons at CP content of 12.8 and 8.7%, respectively. Decline in nutritional quality of Chloris gayana with age in this study has also been reported in similar studies by other authors. [46], [47], [48] and [49] report that there is rapid decline in digestibility and protein content of C.gayana with advance in development. Other authors report that C. gayana is characterized by a particularly low nutritive value of stems compared to leaves and that it is also influenced by the season and variety [50]; [51].
The trend in NDF content due to age at cutting was similar with ADF and ADL and significantly increased (P < 0.05) with advance in maturity confirming the results of similar studies by [52]) and [53]. NDF is relevant to the improvement of the forage nutritional value and can be an important parameter to define the forage quality, because the more fibrous pasture occupies more space for longer and limits the intake rate. For all grasses used in the study, the value of NDF was high during the third harvest (week 28) and varied between 63.3-73.8%. A high NDF that is above 72% will cause low intake of J a n u a r y 08, 2 0 1 6 forage [54] and as NDF percentages increase, dry-matter intake generally will decrease [55] . High NDF content can be attributed to the low development of the grasses in the dry period and to forage maturation [56]. The NDF content for Chloris gayana cv KATR3 was above 72% during all the harvest intervals and this according to [54] can cause low intake. Mulato II maintained lowest NDF content in all the harvest intervals ranging between 56.1% and 63.3%. The B. brizantha cultivars had NDF values ranging between 60.3 and 71.3% of which the highest value of NDF at 71.3% was demonstrated by Basilisk at week 28 harvest. ADF is the value that refers to the cell wall portions of the forage that are made up of cellulose and lignin. These values are important because they relate to the ability of an animal to digest the forage. The digestibility of foods is related to the fiber because the indigestible portion has a proportion of ADF, and the higher the value of ADF the lower the food digestibility [57]. Xaraes, Mulato II and Marandu maintained below 40% ADF throughout the harvest periods. [58] reports that forage with ADF content around 40%, or more, shows low intake and digestibility. Studies conducted by [59] report that intercropping annual crops eg. corn with forages is a good option for providing quality food at critical periods of drought, as from the corn harvest; there is recovery of emergence of new tillers, providing yield and forage with good digestibility. [59] further report that forages grown in integrated systems with corn were still able to attain below 40% ADF at the fourth harvest. Chloris gayana cv. KATR3 maintained higher values for ADF during all the harvest intervals. Acid Detergent Lignin content for all the cultivars increased with harvest interval and was in conformity with other reports by [38], [40] and [60] that showed that ADL content increases with the advance in harvesting days of forage crops. [61] has reported that the soil fertility could also influence grass lignin concentration. Chloris gayana KATR3 recorded higher ADL (5.2-6.5%) content in all the harvest intervals although at week 28 harvest, values for ADL for all the cultivars were similar. Mean ADL values for the grasses ranged between 3.9-4.9% which is within the range reported by [62] that the lignin contents of marginal land grasses at early bloom varied between 2.8% to 4.6% and at mature stage lignin contents ranged between 3.4 to 5.7%.
Decreasing values of DMD may possibly be explained by the advancing physiological maturity, due to the seasonality of forage production which increases the cell wall components and reduces digestibility. The age at cutting of forage crops also has an influence on DMD, which is a function of the chemical constituents of forages [63]. Dry Matter Digestibilty values of the grasses ranged between 43.6-57.5% far less than the range reported by [64] for the digestibility of cultivated tropical grasses which lie between 50 and 65% and of temperate grasses which lie between 65 and 80%. Piata, Basilisk and Mulato II had higher digestibility compared to the other grasses at harvest interval week 28. Values of IVDMD in signal grass of B. decumbens has been found to range between 60% to 70% in immature forage, and from 50% to 60% in mature forage, higher than the average (55%) for tropical forage grasses found by [65]. The DMD values for the Chloris gayana cv KATR3 agree with those reported by [66] of 40 to 60% for sole Rhodes grass. Basilisk recorded digestibility values ranging between 37.2% (at week 22 harvest) to 50.1% (at week 28 harvest interval) which is almost in the range of 42.7 and 50.3%, obtained by [67]. [67], however, reports that the Brachiaria decumbens in his experiment was subject to the shadow of a leguminous tree.
Mean metabolisable energy during the harvest intervals decreased from 6.6-5.9 MJ/kg DM. The highest metabolisable energy values obtained in this experiment were for Xaraes, Marandu and Mulato II of 7.0 MJ/kg DM. In relation to the metabolizable energy, [68] report that pasture and forage with values exceeding 8.37 MJ/kg DM are considered of good quality.

CONCLUSION
The stage of maturity seems to influence forage quality with Crude protein, phosphorus, DMD and ME tending to decrease with advancing maturity. On the other hand NDF, ADF ADL and calcium increase with advancing maturity. The grass cultivars in this study vary as to digestibility and nutritive composition. The Brachiaria species are richer in ash and phosphorus compared to C. gayana KATR3 and Napier grass. Chloris gayana records higher calcium content than all the grasses. MG4 and Mulato II are the only grasses that meet CP requirement for microbial rumen function throughout all the harvest stages. Chloris gayana records high NDF, ADF and ADL and consequently lower digestibility than all the other grasses. The low energy values obtained from this study suggest that suggest that energy may be a limiting nutrient in all these grass cultivars. Brachiaria cultivars however, are superior to both Napier and Chloris gayana in CP, digestibility, and fibre and may be able to supplement these nutrients when intergrated with the indigenous forages to boost livestock feed resources in arid and semi arid regions in Kenya. Further research is needed to establish their animal production potential in these regions.