General information

The current experiment ran from October 2010 to January 2011 at the Semi-arid Experimental Station of the Brazilian Agricultural Research Corporation (EMBRAPA) in the municipality of Petrolina, Pernambuco State, Brazil (09°09′S, 40°22′W, 365 m a.s.l.). The climate is typically semi-arid with an annual rainfall of 570 mm and average maximum and minimum temperatures of 33 and 20 °C, respectively. Low relative humidity and precipitation and high temperatures characterized the experimental period (Table 1).

Table 1.

Meteorological data during the experimental period

Source: Agro meteorological station in the Experimental Area of Bebedouro, Embrapa Semiarid – Petrolina – Pernambuco State, Brazil.

Dietary treatments

Clean water was pumped from EMBRAPA's supply system and stored in the 500-litre storage tanks, which were then isolated with a lid to minimize light and evaporation and to moderate water temperatures. Using this water, treatments were formulated every 2 days by adding sodium chloride (NaCl) until reaching saline concentrations of 640, 3187, 5740 and 8326 mg TDS/l (Table 2), determined and monitored continuously by a portable conductivity meter (Handheld metre 315i model, Weilheim, Oberbayern, Germany). Treatments were then homogenized manually and transferred from the 500-litre tanks to 15-kg water pails, and replenished and metred three times a day (at 08·00, 12·00 and 15·30 h) in order to determine daily drinking water consumption. The saline water concentrations of the current study were selected according to the results of NRC (2001) that reported negative effects of salty water on DMI. Sodium chloride was added to drinking water to increase the TDS concentrations because this element is found in abundance in groundwater and wells of the Brazilian semi-arid region, where the current study was completed (de Lima et al. Reference de Lima, do Nascimento, Guilera, Brandão, Lopes Neto and Feitosa2008).

Table 2.

Electrical conductivity, chemical composition and alkalinity of saline drinking water containing increasing concentrations of total dissolved salts (TDS)

Drinking water intake (in l/day) was measured individually and calculated as saline water offered ad libitum (l/day) to the animals – (saline water orts (l/day) + evaporation (l/day)). Daily water loss by evaporation (in l/day) was measured, since saline water was exposed to evaporation in areas adjoining the metabolism pens, such that the initial measure of saline water contained in pails was subtracted from its final measure after a 24-h period.

Total water intake (in l/day) was also measured individually and calculated as drinking water intake (l/day) + water ingested through the feeds (l/day) per individual animal. About 100 g each of food given and refusals were taken daily from each heifer and oven dried at 65 °C for 72 h to determine water consumption through the feeds.

Retained water (in l/day) was calculated as (total water intake (l/day)) – (water excreted through the faeces (l/day) and urine (l/day)).

Weekly, water samples (250 ml) were collected from each treatment by dipping a sterile bottle into the storage tanks and pails at a depth and place where cattle would normally drink. Then, samples were immediately cooled and delivered to the EMBRAPA Laboratory to measure electrical conductivity, temperature and pH, and to determine the concentrations of HCO_3–, Cl^–, Ca, Mg, K and Na as described by EMBRAPA (1997) (Table 2).

Digestion and performance studies

All the Red Sindhi heifers were cared for in accordance with guidelines of the National Council for the Control of Animal Experimentation (CONCEA 2008), such that the animals were housed, haltered and tethered with enough space for lying, standing, drinking and eating. Twenty-four Red Sindhi heifers (initial body weight (BW) of 200 ± 5 kg) were used to assess the impact of increasing concentrations of TDS (six heifers per treatment) in drinking water on intake, total apparent digestibility, performance, microbial protein synthesis, water and N balance over a 71-day period, where balance describes a calculated total body balance of water and N (‘in’ and ‘out’ flux). Heifers were blocked by weight, assigned randomly to one of the four treatments groups (640, 3187, 5740 and 8326 mg TDS/l) and housed in individual pens (2·0 × 2·60 m). The first 15 days were used to adapt heifers to the diets and pens that were bedded with straw in an open-sided building. Heifers were provided with a total mixed ration (TMR) ad libitum, consisting of (as fed) Buffel (Cenchrus ciliaris) grass hay and concentrate (soy meal, ground maize grain and a supplement containing vitamins and minerals) in a ratio of 50 : 50 (Table 3) and they had free access to water (saline water treatments in pails as described previously) at all times. Rations were prepared twice daily at 07·30 and 16·30 h in a manner that assured 0·10 orts at the morning feeding, with feeds and orts being weighed daily and used to calculate average daily intake as described by da Silva & Leão (Reference da Silva and Leão1979). Apparent digestibility was assessed from day 43 when collections of faeces, orts, feeds and urine were conducted over a 5-day period, during which sub-samples (0·10 of total daily samples) were composited for each animal. To determine total tract digestibility, 500 mg of LIPE™ was infused orally into each animal from day 42 to 47, with apparent digestibility being calculated as concentrations of fed LIPE™ divided by concentrations of excreted LIPE™ using the ratio technique. Faecal samples (c. 100 g fresh weight) were extracted daily from the rectum at 08·00 and 15·30 h to determine total tract digestibility from individual animals using LIPE™ (enriched and purified isolated lignin of Eucalyptus grandis) as an external marker, as outlined by Saliba et al. (Reference Saliba, Rodriguez and Pilo-Veloso2006). An additional faecal sample (c. 100 g fresh weight) was collected from each animal prior to feeding the marker to correct calculations for initial LIPE™ excretion. These faecal samples were composited by animal, dried at 55 °C for 48 h in a forced air oven, ground through a 1-mm screen and sent to the Animal Nutrition Laboratory (Federal University of Minas Gerais, Belo Horizonte, Brazil) to determine LIPE™ concentrations (Saliba et al. Reference Saliba, Rodriguez and Pilo-Veloso2006).

Table 3.

Chemical composition of ingredients and diet

TMR, total mixed ration.

* Contained 190 g/kg calcium, 60 g/kg phosphorus, 30 g/kg magnesium, 35 g/kg potassium, 70 g/kg sodium, 50 g/kg chlorine, 0·02 mg/kg sulphur, 15 mg cobalt/kg, 700 mg copper/kg, 40 mg iodine/kg, 700 mg iron/kg, 1600 mg manganese/kg, 1000 mg selenium/kg, 2500 mg zinc/kg, 200000 international unit (IU) vitamin A/kg, 50000 IU vitamin D/kg and 1500 IU vitamin E/kg.

† NDFap: neutral detergent fibre corrected for residual ash and protein.

‡ ADFap: acid detergent fibre corrected for residual ash and protein.

Spot urine samples (0·20 of total daily output) were also collected over the 5-day collection period from each heifer before morning and afternoon feeding by stimulating urination through vulva massage. The 24-h urinary volume and the purine derivatives (PD) obtained from the spot sampling were estimated from creatinine concentration excreted in urine (Ørskov & MacLeod Reference Ørskov and Macleod1982), because total urine collection was not carried out (Chen et al. Reference Chen, Mejia, Kyle and Ørskov1995; Valadares et al. Reference Valadares, Broderick, Valadares Filho and Clayton1999; Barbosa et al. Reference Barbosa, Valadares, Valadares Filho, Veras, Leao, Detmann, Paulino, Marcondes and Souza2006). Afterwards, urine samples were cooled on ice, transported to the Animal Nutrition Laboratory of EMBRAPA (Petrolina, PE, Brazil) and acidified with 500 ml of 1 M H₂SO₄ to achieve urine pH<3. Aliquots (50 ml) from daily urine samples were composited for each heifer and diluted 1 : 5 with distilled water for analysis of total N and uric acid. Additional sub-samples (25 ml) were also diluted 1 : 50 with distilled water for allantoin analysis. All urine samples were stored at −40 °C for later determinations of PD.

Excretion of microbial PD (MPD) in the urine was estimated using the equation of Verbic et al. (Reference Verbic, Chen, MacLeod and Ørskov1990):

$$\displaystyle{{{\rm MPD} = [{\rm PD} - (0 {\cdot} 385 \times {\rm BW}0 {\cdot} 75)]} \over {0 {\cdot} 85}}$$

where PD (mmol/d) is excretion of the purine derivatives, allantoin and uric acid; BW (kg) is body weight; 0·85 is the recovery of MPD; and 0·385 (mmol/d) is the contribution of endogenous PD. Microbial protein synthesis in the rumen (MPSR, g N/d) was calculated according to Chen et al. (Reference Chen, Chen, Franklin, Orskov and Shand1992):

$$\displaystyle{{{\rm MPSR} = ({\rm MPD} \times 70)} \over {(0 {\cdot} 83 \times 0 {\cdot} 116 \times 1000)}}$$

where 70 is the N content of purines, in mg/mmol; 0·83 is the digestibility of microbial purines; and 0·116 is the ratio of purine N to total N in rumen bacteria.

Microbial crude protein (CP) (g/day) was calculated as MPSR × 6·25.

Whole-body N balance (g/day) was calculated as N intake (g/day) – (N excreted in faeces (g/day) + N excreted in urine (g/day)).

The ADG and feed conversion were measured from day 16 onwards. Heifers were weighed every 28 days and the ADG was calculated over a 56-day period by dividing BW gain (initial full BW − final full BW) by 56. Conversion efficiency was calculated as the ratio between DMI and ADG (kg DMI/kg BW gain).

Chemical analysis

All feeds used, including TMR, orts and faeces composites were dried at 65 °C for 72 h and ground to pass through a 1-mm screening using a Wiley Mill (Tecnal Ltd., São Paulo, São Paulo, Brazil). Ground samples were analysed for dry matter (DM) (method Instituto Nacional de Ciência e Tecnologia – Ciência Animal (INCT-CA): G-003/1), organic matter (method INCT-CA: G-003/1), ash (method INCT-CA: M-001/1), NDICP (neutral detergent insoluble CP) (method INCT-CA: N-004/1) and ADICP (acid detergent insoluble CP) (method INCT-CA: N-005/1) as described by Detmann et al. (Reference Detmann, Souza, Valadares Filho, Queiroz, Berchielli, Saliba, Cabral, Pina, Ladeira and Azevedo2012). A Leco combustion N analyser (FP-428N Determinator, Leco Corporation, St Joseph, MI, USA) was used to measure N content. Crude protein was calculated as N × 6·25. Neutral detergent fibre (NDF) was determined as described by Van Soest et al. (Reference Van Soest, Robertson and Lewis1991) using heat stable α-amylase and sodium sulphite (ash free). The content of acid detergent fibre (ADF), acid detergent lignin (ADL) and residual ash were determined using the sequential method of Van Soest et al. (Reference Van Soest, Robertson and Lewis1991). An Ankom Fibre Analyser (Ankom Technology Corporation, Macedon, NY, USA) was used to extract and filter NDF and ADF. The concentration of hemicellulose was calculated as NDF minus ADF, cellulose content as ADF minus ADL, and lignin concentration as ADL minus residual ash. The ether extract (EE) content was determined as described by AOAC (1995) (method 920·39) using an Ankom Fat Extractor (Ankom Technology Corporation, Macedon, NY, USA). Total (TC) and non-fibrous (NFC) carbohydrates were calculated as described by Sniffen et al. (Reference Sniffen, O'Connor, Van Soest, Fox and Russell1992):

$$\eqalign{{\rm T}{\rm C}_{{\rm g/kg DM}} & = {\rm} 100- ({\rm CP} + {\rm EE} + {\rm ash}) \,\, {\rm and}\,\, {\rm NFC}_{{\rm g/kg DM}} \cr & = 100- ({\rm CP} + {\rm EE} + {\rm ash} + {\rm NDF})}$$

The concentrations of total digestible nutrients (TDN) were calculated according to Weiss & Wyatt (Reference Weiss and Wyatt2000) as:

$$\eqalign{{\rm TD}{\rm N}_{{\rm g/kg DM}} & = {\rm digestible}\,\, {\rm CP} + (2 {\cdot} 25 \times {\rm digestible}\,\, {\rm EE}) \cr & \quad + {\rm digestible}\,\, {\rm NDF} + {\rm digestible}\,\, {\rm NFC}}$$

Urine samples were analysed for total N by the macro-Kjeldahl procedure, uric acid (Human, Itabira, MG, Brazil), creatinine (Human, Itabira, MG, Brazil) and allantoin according to the method described by Chen & Gomes (Reference Chen and Gomes1992).

Statistical methods

Data were analysed as a completely randomized design using the GLM procedure of SAS Version 9.1 statistical program (SAS 2002), with both treatments and heifers being considered as fixed effects. The model used was:

$${\rm Yij} = \mu + {\rm Ti} + {\rm Eij}$$

where Yij is the observation, μ the overall mean, Ti the treatment and Eij the residual error. When significant, concentrations of TDS means were compared using Fisher's protected LSD (i.e. the DIFF option of the LSMEANS statement). Polynomial contrasts were used to determine linear and quadratic effects of TDS inclusion rate. Significance was declared at P < 0·05, and trends were discussed at P < 0·10.