The experimental procedures were carried out at the Ruminant Research Unit, Department of Animal Science, Faculty of Agriculture at Kamphaeng Saen, Kasetsart University, Nakhon Pathom, Thailand, in accordance with approval by the Animal Usage and Ethics Committee of Kasetsart University, Thailand (ACKU62-AGK-007).
Animals, experimental design, and feeding
Thirty-six Holstein crossbred steers (aged 22 months) with an average initial body weight of 453.0±35.3 kg were used in the experiment. All animals were raised in open-air pens (2.5 m × 4 m) and had free access to freshwater. The experimental units were randomly assigned to three different starch sources of concentrates (Table 1): ground corn (GC), ground cassava (CA), or pineapple stem starch (PS) in combination with 2 feeding periods: 1) 1 to 206 days (short-term feeding) and 2) 1 to 344 days (long-term feeding) with six replicates per feed-dietary treatment arrangements as a 2 ´ 3 factorial in a completely randomized design. Some feedstuffs for the starch source (ground corn and ground cassava) were purchased from a commercial feed company (Betagro Co., Ltd., Samut Prakan, Thailand), while the pineapple stem starch was obtained from Hong Mao Biochemicals Co., Ltd., Rayong, Thailand. The pineapple stem starch was dried under sunlight for 2–3 days. The other feedstuffs (rice bran, defatted rice bran, defatted palm kernel meal) were purchased from Cowboyfriend Co., Ltd., Nakhon Pathom, Thailand. For roughage, Napier grass (approximately 65-days old) was harvested from the Ruminant Research Unit, Department of Animal Science, Faculty of Agriculture at Kamphaeng Saen, Kasetsart University, Nakhon Pathom, Thailand, chopped into 2–3 cm lengths, and ensiled. Rice straw was purchased from a dealer in Suphan Buri province, Thailand. The steers were fed the concentrate ad libitum with Napier grass silage (2 kg DM/head/day) and rice straw (0.9 kg DM/head/day). The diets (concentrate and roughage) were offered twice a day at 7:00 am and 4:00 pm. The initial weight, final weight, and dry matter intake (DMI) were recorded and body weight gain (BWG), average daily gain (ADG) and the feed conversion ratio (FCR) were calculated to analyze growth performance.
Feed sampling and analysis
Samples of the concentrate and roughage diets (Napier silage and rice straw) were taken every 2 weeks for nutrient composition analysis. The samples were analyzed for dry matter (DM), crude protein (CP), ether extract (EE), neutral detergent fiber (NDF), acid detergent fiber (ADF), calcium, phosphorus, ash (AOAC 2016), and total carbohydrates (Sniffen et al. 1992). The metabolizable energy (ME) was estimated using ME (Mcal/kg DM) = 1.01 × DE (Mcal/kg) - 0.45 and the digestible energy (DE) was estimated using DE (Mcal/kg DM) = 0.04409 × TDN), according to NRC (2001). Total digestible nutrients (TDN) was calculated using TDN (%DM) = 87.84 – (0.07 × ADF), according to Schmid et al. (1976).
Slaughter, carcass characteristics, muscle sampling, and analysis
When the target day of feeding was reached, the animals were deprived of feed but allowed access to freshwater. According to the halal protocol, animals were fasted for approximately 16 h before being humanely slaughtered (Zainalabidin et al. 2019). After slaughtering, the hot carcasses were weighed and chilled at 4°C. The hot carcass dressing percentage was calculated using (hot carcass weight/slaughtered live weight) × 100. After 7 days of chilling, the cool carcass percentage using (chilled carcass weight/slaughtered live weight) × 100. Muscle samples from the longissimus dorsi muscle (LM) were cut between the 12th and 13th ribs on the righthand side and transported immediately at 4°C to the laboratory for physical analysis, before being frozen prior to further chemical analysis. The pH of the LM was measured using a portable pH meter (TESTO205 pH/temperature meter; Testo Pty Ltd., Croydon South, VIC, Australia).
The drip loss of the LM was evaluated by the weight difference pre-hanging and post-hanging of the meat samples at 4°C for 24 h (Honikel 1998). Meat color (L* = lightness; a* = redness; b* = yellowness) was measured on a freshly cut, transverse surface after 30 min of blooming at three spots using a color meter (HunterLab Mini Scan EZ, Reston, VA, USA) according to the protocol of CIE (1978). The irregular shape of the LM area was measured using a compensating planimeter (Cacere et al. 2014). Backfat thickness was measured using a set of calipers between the 12th and 13th ribs on the LM at three-quarters of the length of the loin eye muscle from the chine (backbone) (Orellana et al. 2009).
Chemical and fatty acid composition in meat analysis
The moisture, crude protein, and crude fat contents of the LM were determined in duplicate (AOAC 2016). The fatty acid profile of IMF was extracted using the direct fatty acid methyl ester synthesis method (Folch et al. 1957) and determined using a gas capillary chromatograph on a CP 7488, 0.20 mm ´ 50 m, CP – Sill 88 capillary column (Chrompack CP 9001 chromatograph, Netherlands). The initial oven temperature was 140 °C at 4 °C min-1, held for 10 min, then to 200 °C at 10 °C min-1, and held for 13 min. Helium was used as the carrier gas at a flow rate of 100 kPa. The injector was set at 270 °C and detector at 280 °C. Fatty acid profiles were identified by comparing their retention times with those of commercially available reference standards purchased from Sigma Aldrich (Germany).
The production cost was determined by considering the costs of the steers, feed, and the total cost of each animal according to the methods of Boonsaen et al. (2017) and Pintadis et al. (2020). The fixed costs included depreciation price of land, building and equipment, and the variable costs included wages, medicine, water, electricity, fuel, materials, and interest.
The economic analysis used the average regional prices for 2020. The key factors were the chilled carcass weight at 7 days post-modern and the marbling grade (Beef Cluster Cooperative Limited, Nakhon Pathom, Thailand). The price depended on the marbling score: score 1 = devoid at USD 5.74 per kg chilled carcass, score 2 = slight at USD 6.39 per kg chilled carcass, score 3 = small at USD 6.89 per kg chilled carcass, score 4 = moderate at USD 7.21 per kg chilled carcass, and score 5 = abundant at USD 7.54 per kg chilled carcass, where the conversion rate used was USD 1 = THB 30.50.
Based on the above valuation, the profit per head of steers raised during the finishing period (short term period or long-term period) was calculated using: Profit per steer = Total revenue per steer - Total cost per steer.
Growth performance, carcass characteristics, and meat quality were analyzed as a 2 ´ 3 factorial in a completely randomized design, using the following model: Yijk = μ + Fj + Si + FSij eijk, where Yijk is the observed value of the dependent variable, μ is the overall mean, Fj is the effect of the days on feed (j = 1 to 2; 1 = short-term feeding, 2 = long-term feeding), Si is the effect of the treatment i (i = 1 to 3; 1 = GC, 2 = CA, 3 = PS), and eijk is the experimental error. The hot carcass and cool carcass percentages, the crude fat, crude protein contents, the drip loss, and the fatty acid profile were transformed using an arcsine square-root transformation (Cook 1999) to achieve a normal distribution of the experimental error. Analysis of variance (ANOVA) was used. Tukey’s honest significant difference test was conducted for multiple comparisons post-hoc. All statistical analyses were conducted using the R software, Version 3.6.2 (R Core Team 2020). Significance was tested at P < 0.05.