Introduction

Population growth and shrinking of grazing land have compelled farmers to shift from extensive to intensive dairy system in order to optimise milk yield per cow (Lukuyu et al. 2012). In spite of the additional stress on limited feed resources, especially during the dry season, farmers retain female calves to replace culled cows (Mohd Nor et al. 2015). In tropical areas of Asia, Africa, and South American highlands, farmers lose replacement dairy stock due to limited knowledge on calf and heifer rearing. Approximately 35 % of the losses can be restored using adequate feeding (Moran 2011). In these areas, Napier grass (Pennisetum purpureum) is the most abundant single, year-round feed resource in smallholder dairy farms (Rahman et al. 2015; Mutimura et al. 2013). However, total dependence of farmers on Napier grass is risky because of Napier stunt disease that poses threats to the production of this grass throughout the East African region (Asudi et al. 2015). Developing disease resistant cultivars has been identified as one possible approach to address the problem (Kawube et al. 2014). However, there is need to consider alternative fodder species to complement the search for disease resistance in the global germplasm collection and local landraces.

Brachiaria species are indigenous grasses to Africa, which have been selected for productivity and tolerance to abiotic and biotic stresses in Latin America (Miles et al. 2004). Brachairia hybrid cultivar (cv.) Mulato II was introduced, evaluated, and selected by farmers in Rwanda (Mutimura and Everson 2012). However, its superiority over Napier grass in terms of animal productivity in stall-fed cattle has not been examined. Data on animal growth performance from different Brachiaria grass species is limited to grazing trials (Gracindo et al. 2014). The objective of this study was to evaluate the relative intake and growth performance of crossbred dairy heifers fed on Brachiaria hybrid cv. Mulato II compared with Napier grass under a cut-and-carry forage feeding system in Rwanda.

Materials and methods

Location

The feeding trial was conducted at Songa research station of Rwanda Agriculture Board (RAB). The station is located in the mid-altitude zone (1471 m.a.s.l) of Rwanda, and it lies between 29° 48′ E, 2° 25′ S. The average annual rainfall is 1087 mm and relative humidity of 77 % with an average temperature of 20.1 °C per year.

Management of animals

Sixteen (Ankole × Jersey) crossbred heifers (605 ± 11 days of age and 203 ± 35 kg live weight) were selected and divided into two groups of eight animals. The animals from each group were ear tagged, randomly assigned to one of the two dietary treatments. Animals were put in individual pens in a house built for cows in the station and partitioned for stall feeding. ALBENDOZOLE (10 mL/10 kg body weight) and acaricide (Norotraz 12.5 % effective concentration (EC), 2 mL/1 L of water; twice/week) were used to control endo and ecto-parasites, respectively. Individual pens were cleaned every morning.

Feeds and feeding

The dietary treatments were two different roughages: Brachiaria grass (Brachairia hybrid cv. Mulato II) or Napier grass (Pennisetum purpureum) fed as basal diets. All animals received commercial concentrate supplements (2 kg/day) which was composed of maize (55 %), soybean (10 %), rice bran (10 %), palm cakes (20 %), bone powders (1.5 %), salt (0.5 %), and molasses (3 %). Water and mineral blocks were provided ad libitum. The basal feeds (grasses) were harvested (15 cm above ground) from the station plots where they were planted without fertiliser application. The soil type of the plots is sandy clay with nitrogen and carbon content of 0.2 ± 0.4 % and 1.2 ± 0.5 %, respectively. The harvested herbages were chopped (10 cm length) using forage chopper (Mild steel, 7 HP of power, electric motor/diesel engine, BrazAfric Ltd) before feeding. Basal diets were given at ad libitum based on individual body weights. After an adaptation period of 14 days, daily feed offers and refusals, respectively, were weighed, recorded, and sampled at 9:00 h and 15:00 h for a period of 12 weeks (from 21st February to 21st May 2014). Fortnightly, individual animals were measured to the nearest 100 g using mechanical weigh bridge (PORTEE 1000 kg, 2 × 1 m, B.C, 188021, RAPPORT).

Daily feed dry matter (DM), organic matter (OM), crude protein (CP), metabolisable energy (ME), calcium (Ca), and phosphorus (P) intake were calculated as the difference between feed offer and refusal corrected for the respective contents in the original samples (Balehegn et al. 2014). Feed conversion ratio (FCR) was calculated as the slope of the linear regressions of cumulative nutrient (DM, OM, and CP) intakes on growth rates. Daily ME requirement for growing heifers was calculated based on Eqs. 1, 2, 3, 4, and 5 (AFRC 1993). Growth rates (g/day) were estimated as the slope of the linear regressions of weekly body weights on days of feeding.

$$ M{E}_m\kern0.5em \left(MJ/\mathrm{day}\right)=\left(F+A\right)/{k}_m $$
(1)

where ME m is the metabolisable energy for maintenance; F is the fasting metabolism; A is the activity allowance, and K m is the efficiencies of utilization of metabolisable energy for maintenance which is 0.665 for a heifer (AFRC 1993).

$$ F\kern0.5em \left(MJ/\mathrm{day}\right)={C}_1\left\{0.53{\left(W/1.08\right)}^{0.67}\right\} $$
(2)

where W is the body weight and C 1 is the Constant which takes 1 for other cattle except bull.

$$ A\kern0.5em \left(MJ/\mathrm{day}\right)=0.0071W $$
(3)
$$ E{V}_g\kern0.5em \left(MJ/\mathrm{kg}\kern0.5em WG\right)=\frac{C_2\left(4.1+0.0332W-0.000009{W}^2\right)}{\left(1-{C}_3\times 0.1475\varDelta W\right)} $$
(4)

where EV g is the energy values for growth; WG is the weight gain; C 2 is the corrects for body size and sex, for heifer at early mature, it is 1.3; C 3 = 1 when it is a plane nutrition; ∆W is the body weight change.

Energy retained in animal’s body per day (E g ) was calculated as follows:

$$ {E}_g\kern0.5em \left(MJ/\mathrm{day}\right)=\varDelta W\times E{V}_g $$
(5)

Chemical composition of feeds used

Samples of feed offered and refusal were collected daily. Samples of every week were mixed and two samples were taken and analysed for chemical composition. The official protocol was used to the determine DM, Ash, and OM (AOAC 1990; method ID 9420.5) and CP (AOAC 2006; method ID 984.13). Macro and micronutrients were determined using Atomic Absorption and Flame Emission Spectrophotometer (PerkinElmer, Inc., Precisely, A. Analyst 200).

Data analysis

Chemical compositions of feeds over 12 weeks were analysed using general linear model (GLM) procedures of the Statistical Analysis System (SAS 2010). Means were compared using PDIFF option of SAS. Data from experiments on feed intake and live weight gain were subjected to analysis of variance (ANOVA) in a completely randomised design using GLM procedures of the Statistical Analysis System (SAS 2010) based on the following model (Eq. 6):

$$ {Y}_{ijk}=\mu +{H}_i+{F}_j+{e}_{ijk} $$
(6)

where Y ijk is the variable dependent; μ is the boverall mean; H i is the animal effect; F j is the effect of feed; and e ijk is the residual error.

Initial body weight of heifers was used as a covariate in analysis of the effect of diets on body weight gain. Individual and group animal differences between means were separated using least significance difference (LSD) at P < 0.05 level of significance.

Results

Feed composition

Chemical composition of the feeds used in this experiment is presented in Table 1. Ash contents of the roughages were significantly higher than the contents in concentrates (P < 0.05). Brachiaria hybrid cv. Mulato II had less ash content than Napier grass (P < 0.05). Commercial concentrates had more OM and CP than roughages (P < 0.05). OM and CP in Napier grass were lower than those of Mulato II (P < 0.05). The roughages and concentrates did not differ in Ca contents (P > 0.05) but the roughages had lower contents of P than the concentrates (P < 0.05; Table 1).

Table 1 Chemical composition of feed used in the experiment
Full size table

Intake

Absolute (kg or g/day) and relative (kg or g/kg metabolic body weight, BW0.75) daily intake of DM, OM, CP, and Ca were significantly (P < 0.01; Table 2) higher in animals fed Mulato II supplemented with CC (MCC) than Napier grass supplemented with concentrates (NCC) as basal diets. However, P intake was higher in NCC than in MCC diets.

Table 2 Effect of roughage on intake of feeds and its nutrients by crossbred heifers
Full size table

Body weight gain and feed conversion ratio

Results from live weight gain (LWG) and feed conversion ratio (FCR) are shown in Table 3. The final body weight (FBW) and average body weight gain (ABWG) were similar (P > 0.05) between the two roughages. Although there was no significance difference (P > 0.05) between dietary groups, average daily weight gain (ADWG) of heifers fed on MCC diet was numerically higher than those fed on NCC diet. Feed conversion ratio (FCR) for DM, OM, and CP was significantly different (P < 0.001) between MCC and NCC diet. This suggests that high numerical body weight gain observed in MCC diet was due to higher FCR.

Table 3 Bod weight gain and feed conversion ratio of crossbred dairy heifers fed on MCC in comparison to NCC diet
Full size table

Discussion

Dry matter intake (DMI) and contents of nutrients in feeds are major factors determining feed quality and animal productivity (McDonald et al. 2011). In the present study, we found that Mulato II was better than Napier grass as a potential source of protein and energy. Although diets offered to crossbred dairy heifers differed in CP and OM, no variation in P and Ca was observed. Higher values of OM and CP in Mulato II than in Napier grass were reported in previous studies (Mutimura et al. 2015). DM and nutrient intakes were higher in MCC diet than in NCC diet. In this respect, Mulato II had comparative advantage in DM intake than Napier grass because of its leafiness and thinner stems than Napier grass (Maass et al. 2015). Therefore, the animals could eat more Mulato II than Napier grass. Also, high DMI in MCC diet might have influenced by high CP content in the diet. This observation is in agreement with Malisetty et al. (2014) who reported that DMI increases with the increase of CP content in a diet. Morais et al. (2014) also reported that when quality of supplement and supplementation frequency remain the same, the difference in weight gains of an animal will be based on the quality of roughage. As the two groups of crossbred dairy heifers had received the same amount of the commercial concentrates, the major factor which influenced differences in DMI might have been the quality of roughages where MCC had higher CP and MO content than NCC (Table 1).

High-crude protein CP intake was observed in MCC diet, and this diet had high CP. This suggests that CP content in feed influenced its intake. This agrees with Singh et al. (2015) who reported increase of CP intake when CP was increased in a feed. CP intake of 0.8 kg/day was slightly higher than results reported on CP intake from corn meal supplemented with jatropha and fed on Holstein heifers (da Silva et al. 2015). However, our results were higher than those reported in a feeding trial when Tho-tho male cattle were fed on tree leaves based ration (Das et al. 2011). Relative DM and nutrient intakes were higher in MCC than in NCC diet (Table 2). Similar findings were reported by Ngim et al. (2011) and suggested that grass with high relative intake should be integrated in livestock feeding system. Generally, the trend showed that diet with high nutrient content had higher intake of these nutrients; however, this trend was different for minerals. This is because both diets had similar P content but higher P intake was observed in NCC diet. Although the explanation of this observation seems complicated, however, previous studies have reported similar trend where Ca and P intakes did not correlate with their concentration in a diet (Sinha et al. 2011).

Body weight changes from the two groups were not statistically different, but numerically average daily weight gain (ADWG) of heifers fed on MCC exceeded those fed on NCC. Ngim et al. (2011) reported similar results on cattle fed on Mulato II as the basal feed in comparison with other grass in Thailand. In addition, differences in CP, OM, and ME intakes between the two dietary groups are attributable to increased ADWG in MCC diet. ME intake was much higher than requirement for MCC heifers compared to NCC heifers. A positive weak correlation between ME intake and ME requirements was observed in MCC whereas there was no correlation in NCC diet (Fig. 1). This means that high ME consumed was not translated into superior growth performance in heifers fed on NCC diet. However, our results on ADWG of heifers fed on MCC were slightly higher than those reported on crossbred (Friesian × Boran) heifers (532 g/day) and on Bhadawari buffalo heifers (330 g/day) fed on hay and wheat straw supplemented with commercial concentrates, respectively (Singh et al. 2015; Gojjam et al. 2011). Furthermore, FCR values were different between the two dietary groups of heifers. FCR showed that for the heifer to gain 1 kg of live weight per day it should eat 9.3 and 11.5 kg of DM in MCC and NCC diets, respectively. Similar value for FCR (9.5 kg of DM/kg ADWG) was reported when steers were grazing on smooth bromegrass (Lardner et al. 2015).

Fig. 1
figure 1

Relationship between metabolisable energy (ME) intake and metabolisable energy requirement (MEreq) for the two groups of growing heifers (MCC—Y = 0.33x + 40, R 2 = 0.30; NCC—Y = 0.11x + 39, R 2 = 0.03)

Full size image

FCR data showed that diet with low CP and ME had poor FCR. A similar observation was reported when cows were fed on low and high level of protein (Fiems et al. 2015; Wang et al. 2014). It has been also reported that good FCR value was influenced by environment, feed type, and high energy intake (Singh et al. 2015; Fiaz et al. 2012). This suggest that diets should be selected based on their quantity and quality.

Daily body weight gain of heifers fed on MCC diet showed no statistically significant difference but numerically exceeded those fed on NCC diet. Considering DM and nutrient intakes as well as the quality attributes of Mulato II, this forage grass can be integrated into cut-and-carry feeding system in smallholder farms to feed heifers predestined for dairy mature cow replacement.