Behavioural and physiological responses of individually housed dairy calves to change in milk feeding frequency at different ages

Animals

Forty-eight Holstein Friesian calves (22 females, 26 males) born between 5 September 2014 and 14 October 2014 were allocated to the study following weighing at <12 h of age, which was used as the birth weight (41.4 ± 4.7 kg). All calves received 3 L colostrum from their own dam via oesophageal feeding tube or teated bottle within 3 h of birth with a second feed of 3 L colostrum provided 12 h after calving. All calves were provided with transition milk collected from the first four milkings from their own dam on days 2–3 and a mix of transition milk and MR on day 4 as detailed in Table 1. On entering the rearing accommodation at ≤12 h of age, calves were individually housed, with 12 individual straw-bedded pens (0.9 m × 1.8 m) constructed inside each of four large concrete walled housing blocks (6 m × 6 m) using hurdles. Calves remained in individual pens until the end of the experiment at 63 d of age. All housing blocks were contained within the same calf-rearing accommodation. Calves were housed in order of date of birth with four calves from each treatment within each housing block. Calves had visual contact with all conspecifics within each housing block and nose-to-nose contact with calves in neighbouring individual pens. Calves were vaccinated with Bovilis Bovipast RSP (MSD Animal Health, Milton Keynes, UK) at 2 wk of age with a secondary dose given 4 wk later at which point calves were vaccinated with Bovilis IBR Marker Live (MSD Animal Health, Milton Keynes, UK).

Treatments and experimental design

The study commenced when calves reached 5 d of age from which point they were offered MR according to one of three feeding regimes: (i) once-a-day feeding commencing at age 14 d (OAD14), (ii) once-a-day feeding commencing at age 28 d (OAD28) and (iii) twice-a-day feeding (TAD). It was decided to reduce MR feeding frequency at these time points so as to reflect the age at which it has previously been indicated as a stressor (OAD28; Hulbert et al., 2011) and an age at which it has previously been reported as occurring on farm (OAD14; Kehoe et al., 2007). During the experimental period, the liquid diet consisted of a whey-based MR (Volac International Ltd, Hertfordshire, UK). Milk replacer was diluted at a rate of 17.8% and 12.5%, respectively, for calves fed once and twice per day (Table 1). Starter concentrate and fresh drinking water were provided on an ad libitum basis from 5 d of age. Calves were balanced across treatments for sex and birth weight. All calves were abruptly weaned at 56 d of age, and the study was completed when calves reached 63 d of age.

Data collection

Blood sampling

Single-point blood samples were taken via jugular venipuncture using 10 mL clot-activated serum separation Vacutainer tubes (BD, Plymouth, UK) at days −7, −3, 0, +3 and +7 relative to the introduction of the imposed nutritional stressors, being reduction of milk feeding frequency to once daily at 14 or 28 d of age for OAD14 or OAD28 treatment, respectively, and abrupt weaning of all calves at 56 d of age (Figure 1). Samples taken between days 7 and 21 were classified as Period 1, days 21 and 35 as Period 2 and days 49 and 63 as Period 3. Samples were taken 3–4 h post AM MR feed (Kehoe et al., 2007) to allow for analysis of non-esterified fatty acids (NEFA), urea, β-hydroxybutyrate (BHBA) and haptoglobin (Hp). A second sample taken on these days using fluoride oxalate-coated vacutainers (BD, Plymouth, UK) was used for glucose analysis. Blood tubes were centrifuged at 3,000 rpm for 15 min (Eppendorf 5810R, Eppendorf UK, Stevenage, UK) after which the supernatant was decanted and stored at −20°C until analysis by enzymatic assay (Audit Diagnostics Sapphire 800 automated dry chemistry analyser, Audit Diagnostics, Cork, Ireland).

Figure 1.

Timeline of haematological sampling throughout Periods 1–3. Arrow indicates days that nutritional stressors were imposed; commencement of once-daily MR feeding at days 14 and 28 for OAD14 and OAD28 calves, respectively, and abrupt weaning of all calves at day 56. MR = milk replacer, OAD14 = once-a-day feeding milk replacer diet from 14 d of age, OAD28 = once-a-day feeding milk replacer diet from 28 d of age.

Statistical analysis

All data were analysed using GenStat^ (version 16.2; VSN International Ltd, Hemel Hempsted, UK). All statistical models included birth weight as a covariate and housing block and calf as random terms unless otherwise stated. A probability of P < 0.05 was selected as the level of significance, and where data was significant, it was subjected to Fisher’s protected least significant difference test.

Intakes

Starter concentrate intake tended to be increased in OAD14 calves between days 28 and 55 when compared with TAD calves (Table 3; P = 0.073). Both OAD14 and OAD28 calves tended to have increased daily concentrate consumption between days 5 and 55 compared with TAD calves (Figure 2A and Table 3; P = 0.052). There was no effect of treatment on daily concentrate intake or total concentrate intake in the period immediately post weaning (Table 3). Consumption of drinking water tended to be increased in OAD14 calves between days 14 and 27 (Table 3; P = 0.051). Both OAD14 and OAD28 calves consumed more drinking water on a daily basis than TAD calves between days 28 and 55 (Table 3; P < 0.001) and when analysed over days 5–55 (Figure 2B and Table 3). Total water consumption, the cumulative total of drinking water plus water used in reconstituted milk powder, was lower in OAD14 calves between days 14 and 27 (Table 3; P < 0.001) and in OAD14 and OAD28 calves between days 28 and 55 (Table 3; P < 0.001). When analysed over days 5–55, daily total water intake was increased by 287 and 641 mL in TAD calves when compared with that in OAD28 and OAD14 calves, respectively (Figure 2C and Table 3; P < 0.001). There were no treatment differences in drinking water consumption in the post-wean period (Table 3).

Figure 2.

Daily concentrate (A), drinking water (B), and total water (C) intake of calves throughout the pre-wean period. OAD14 (– –) = once-a-day feeding milk replacer diet from 14 d of age; OAD28 (—) = once-a-day feeding milk replacer diet from 28 d of age; TAD (·····) = twice-a-day feeding milk replacer diet.

Figure 3.

Maximum infrared eye temperature of calves over the milk step-down and weaning periods. OAD14 (– –) = once-a-day feeding milk replacer diet from 14 d of age, OAD28 (—) = once-a-day feeding milk replacer diet from 28 d of age and TAD (·····) = twice-a-day feeding milk replacer diet.

Live weight, live-weight gain and FCE

Live weight at weaning (day 56), end of experiment (day 63) and live-weight gains throughout the experimental period for the three treatments are reported in Table 4. Live weight and live-weight gain were unaffected by treatment throughout the experimental period. FCE (kg gain/kg dry matter intake [DMI]) was comparable across treatments between days 5 and 55 (Table 4). There was a tendency for improved FCE in OAD28 calves between days 14 and 27 (Table 4; P = 0.055); however, this did not persist between days 28 and 55 when FCE tended to be improved in TAD calves when compared with OAD14 calves (Table 4; P = 0.076).

Haematological variables

In Period 1, there were no effects of treatment, sampling day or treatment × sampling day on the concentration of blood metabolite variables (Table 5). However, average concentration of Hp increased from 0.30 to 0.40 g/L between days 14 and 21 across all treatments (P < 0.001).

In Period 2, concentration of BHBA increased from 0.12 to 0.17 mmol/L between days 28 and 35 across all treatments (P < 0.001). During the same period, average glucose concentration was lower in TAD calves than in OAD14 and OAD28 calves (Table 5; P = 0.020) and average NEFA concentration was lower in OAD14 calves than in TAD and OAD28 calves (Table 5; P = 0.009). There were no effects of treatment, sampling day or treatment × sampling day on concentration of urea or Hp throughout Period 2 (Table 5).

In Period 3, BHBA concentration peaked on day 59 with an average value of 0.41 mmol/L across all treatments (P = 0.006). In the same period, urea concentration was lower on day 56 (day of abrupt weaning) when compared with days 59 and 63 (P < 0.001). There were no effects of treatment, sampling day or treatment × sampling day on concentration of NEFA, glucose or Hp throughout Period 3 (Table 5).

Thermal imaging

There was no effect of treatment or treatment × age on maximum daily IR eye temperature, IR rectal temperature or manual rectal temperature throughout the three experimental periods. However, during Period 3, there was an effect of age, with maximum eye temperature decreasing as calves aged (Figure 3; P < 0.001). A significant interaction between age and method of taking temperature was found over the course of the experimental period (P < 0.001).

There was no correlation between IR temperature of the eye and manual rectal temperature throughout Period 1 or 2; when all three time periods were combined, however, there was a strong positive correlation between the two methods in Period 3 (R ² = 0.61; P < 0.001). Manual rectal temperature was consistently higher than IR eye temperature across all time periods (P < 0.001). Correlation between IR rectal temperature and manual rectal temperature increased throughout the three experimental periods, with a moderate correlation found between the two methods in Period 3 (R ² = 0.57; P = 0.001) and a low correlation found when all experimental periods were combined (R ² = 0.16; P = 0.004).

Lying behaviour

There were no effects of treatment, age or treatment × age on number of daily lying bouts, lying bout duration or total daily lying time throughout Periods 1 and 2 with calves across all treatments spending approximately 1,122 min per day lying (Table 6). During Period 3, lying time was an average of 48 min less in TAD calves than in OAD28 calves (Table 6; P = 0.043).

There was no difference in hourly lying behaviour between treatments following reduction in MR feeding frequency at 14 d of age (Figure 4; P = 0.183). Hourly lying time tended to be increased in OAD28 calves following reduction in MR feeding frequency at 28 d of age (Figure 4; P = 0.065); however, this difference was minimal, representing an increase of only 1.7 and 1.4 min when compared with that in OAD14 and TAD calves, respectively. There was no difference in hourly lying time following abrupt weaning at 56 d of age (Figure 4; P = 0.325).

Figure 4.

Hourly lying behaviour of calves over the milk step-down and weaning periods. OAD14 (– –) = once-a-day feeding milk replacer diet from 14 d of age, OAD28 (—) = once-a-day feeding milk replacer diet from 28 d of age, TAD (·····) = twice-a-day feeding milk replacer diet.

Heart rate and heart rate variability

There was no difference in HR (bpm) between treatments following reduction in milk feeding frequency at 14 or 28 d of age or following abrupt weaning at 56 d of age (Table 7). During Period 1, a tendency for lower RMSSD (ms), a measurement based on the time between each heartbeat and linked to vagal tone, was found in OAD14 calves on days 14 and 16 when compared with that in OAD28 calves (Table 7; P = 0.061). There were no effects of treatment, sampling day or treatment × sampling day on RMSSD during periods 2 and 3 (Table 7). There was a strong correlation between HR and RMSSD over the experimental period, with RMSSD decreasing as HR increased (R ² = 0.75; P < 0.001).

Intake and performance

The neonatal rumen is underdeveloped, with calves relying primarily on milk or MR as a nutrient source for the first few weeks of life, only beginning to consume measurable amounts of solid feed at around 2 wk of age (Khan et al., 2008). Initiation of solid feed consumption stimulates metabolic and physical development of the rumen (Baldwin et al., 2004; Khan et al., 2011), with research indicating that provision of even a small amount of solid feed to veal calves results in demonstrable progression of forestomach development (Cozzi et al., 2002). An often-overlooked aspect of pre-wean calf feeding management is provision of drinking water. For example, Gottardo et al. (2002) reported that veal calves consumed more than 5 L/d drinking water in addition to the average of approx. 14 L/d of water given with MR. In the same study, calves offered drinking water and straw showed improved rumen papillae growth when compared with calves offered straw and no water. As highlighted by Kertz et al. (1984), sufficient quantities of water contained within milk cannot reach the rumen due to the oesophageal groove closure. Moisture is a limiting factor of gastrointestinal microbial growth in herbivores, meaning that provision of drinking water could help facilitate carbohydrate fermentation and volatile fatty acid (VFA) production; particularly when feeding a low digestible substrate, this stimulates rumen papillae development (Gottardo et al., 2002).

In the present study, OAD calves consumed approx. 500 mL/d more drinking water than TAD calves during the pre-wean period. Jasper et al. (2008) and Budzynska & Weary (2008) reported that calves offered warm water through the automatic milk feeding system for the first 2 d following weaning appeared to substitute milk with water. This resulted in a reduced behavioural response to weaning as indicated by fewer vocalisations when compared with abruptly weaned calves. As indicated by Allen (1996), satiety is partly mediated by gut fill, meaning that the warm water may have acted as a replacement for a proportion of the gut fill previously supplied by the milk feed and thus limited response to weaning. As such, the increased volume of drinking water consumed by OAD calves in the present study was likely a compensatory measure for reduced gut fill as a result of a lower daily MR volume. Burgstaller et al. (2017) reported that the average abomasal emptying time for 2 L of a milk protein-based MR is between 3.2 and 3.4 h. Although increasing MR concentration following the reduction of feeding frequency likely slowed abomasal emptying (Burgstaller et al., 2017), the increase in drinking water consumption in OAD calves could suggest that calves did not reach satiety. This was further compounded by the fact that total volume of water consumed, and therefore likely gut fill, remained below that of TAD calves. It must be considered therefore that feeding low levels of MR once daily represents a welfare concern. Future research should examine both volume and diurnal pattern of drinking water consumption relative to milk feeding frequency and volume as this may be a helpful indicator of calf hunger. Additionally, assessing the effects of MR concentration on abomasal emptying, particularly in once-daily feeding systems, is of importance as this could help to improve feeding management regimes and thus calf welfare.

In the present study, weaning weight and live-weight gain were comparable across treatments and similar to calves fed whole milk once or twice daily and weaned at 56 d of age as reported by Gleeson et al. (2007). FCE of young calves is approximately 50% (Bach & Ahedo, 2008). Recent research has reported no differences in FCE in calves offered high (8 L/d) or low (4 L/d) levels of MR (De Paula et al., 2017) despite differences in DMI from concentrate feed. In the present study, FCE was 0.6 kg gain/kg DMI across all treatments between days 5 and 55, indicating that pre-weaning efficiency was not significantly impacted by feeding frequency. This suggests that if the quantity of MR as solids is the same and health status is high, as in the current study, calves receiving a reduced volume of MR in one feed per day can achieve comparable growth rates and feed efficiencies as those fed an increased volume twice per day. However, although there was no negative impact on calf performance, it must be remembered that affective state of calves was not measured within the parameters of this study. It is therefore possible that calves could have experienced hunger or disruption to behavioural needs as a result of the once-daily feeding regime imposed.

Blood metabolites

Concentration of BHBA has previously been indicated as a potential marker for starter feed consumption and rumen development in young calves (Quigley et al., 1991). The increasing concentrations of plasma BHBA up to the point of weaning in the present study were comparable to the pattern reported by Ferreira & Bittar (2011) and followed the steady increase in concentrate intake. Following weaning, concentrate intake increased rapidly in calves across all treatments between days 56 and 59 resulting in increased BHBA concentration. Previous research reported increased BHBA concentration in gradually weaned calves (e.g. Khan et al., 2007) and was considered to be indicative of enhanced rumen function. In the present study, the tendency for increased BHBA in OAD calves compared with TAD calves at day 59 could therefore potentially indicate the start of earlier ruminal fermentation in these calves. Increased urea concentration in the post-wean period was also suggestive of increased ruminal fermentation of dietary protein as a result of enhanced solid feed intake following the withdrawal of MR (Quigley et al., 2006; Khan et al., 2007). The reduction in BHBA concentration observed across treatments at day 63 is likely as a result of rate of concentrate consumption becoming more stable (Quigley, 1996).

Glucose concentration was comparable to results reported by Ferreira & Bittar (2011), with concentration declining as calf age increased. Lower glucose concentrations in TAD calves during Period 2 were likely due to the increased amount of dietary lactose received in the AM MR feed by OAD calves, with similar effects observed in calves fed on an elevated plane of MR (e.g. MacPherson et al., 2016). The increased glucose concentration in OAD calves during this period could also in some part be due to their increased concentrate consumption, resulting in an augmented DMI (Webb et al., 1969). The lower glucose concentration across all treatments post weaning was likely as a result of the shift from glucose to VFA as the main energy source and the establishment of ruminal function by the calf (Hammon et al., 2002).

Quigley (1996) reported a negative correlation between NEFA concentration and calf starter intake, which could explain the decrease in NEFA concentration in OAD14 calves during Period 2. Increase in NEFA concentration can be indicative of body fat mobilisation due to nutritional stress (Khan et al., 2007); however, no increases were observed following weaning. Therefore, it could be suggested that calves, regardless of milk feeding regime, were consuming sufficient concentrate to support physiological requirements (Ferreira & Bittar, 2011). The combination of higher BHBA and urea concentration coupled with decreasing glucose and NEFA concentrations can be ascribed to the beginning of ruminal fermentation and thus the commencement of transition to functional ruminant (Khan et al., 2007).

Behavioural and physiological measures

Observational and acoustic monitoring techniques are often used to assess calf reaction to common husbandry practices; however, these methods can be time consuming and costly (Rushen et al., 2012). Increasingly, automatic monitoring technologies that allow the non-invasive collection of behavioural and physiological data are being employed on-farm (Sutherland et al., 2018). Previous research investigating the effects of feeding regime on calf behaviour have employed the use of automatic feeding systems, with increases in unrewarded visits to the feeder linked to calf hunger (Rosenberger et al., 2017). However, these are not applicable in systems where calves are individually penned. As such, investigating the use of monitoring technologies that can be used across housing systems is of great importance. Therefore, the main aim of this study was to use novel sensor technologies combined with traditional physiological measures as a means of further elucidating the impact of nutritional change on pre-weaned dairy calves.

Haematological variables

Acute-phase proteins such as Hp have been shown to change in concentration as a response to internal or external challenges caused by stress or infection (Murata et al., 2004). The increase in Hp concentration between 14 and 21 d of age in the present study was possibly as a result of calves receiving primary vaccination at 2 wk of age, which can result in an increase in symptoms of illness. The lack of measurable increase in Hp concentration as a result of reduction in milk feeding frequency at 28 d of age in the present study was comparable to that reported by Hulbert et al. (2011). Although there was no increase in Hp concentration following weaning, as was reported by Kim et al. (2011), the level observed in the present study could again be as a result of the effects of secondary vaccination. The lack of measurable response as a result of reduction in MR feeding frequency or weaning could suggest that these events did not evoke a physiological response measurable by Hp concentration alone.

Infrared thermography

IR thermography has been shown to be a viable method of detecting fluctuations in radiated temperature occurring as a result of physiological changes caused by external stressors (Stewart et al., 2005). Calves did not elicit a measurable response in terms of IR eye temperature change as a result of reduction in milk feeding frequency from twice to once daily; however, IR temperature decreased in the period following weaning. Weaning is regarded as one of the most significant stressors in early life (Weary et al., 2008), and as such, this decrease could potentially indicate a more prolonged stress response to withdrawal of the milk feed. Previous research has indicated that mature cows undergoing periods of feed restriction display a reduction in heat production (Freetly et al., 2006); thus, it is also possible that the decrease in radiated temperature is an acute physiological response to the change in primary energy source as a result of the removal of the milk feed at weaning.

To the best of the authors’ knowledge, neither the variation in the strength of relationship between age and temperature method nor the decrease in IR temperature following weaning has been reported in any other studies. Further research is required to examine effects of age on IR temperature and to investigate the relationship between IR temperature and core body temperature.

Calf lying behaviour

Calf response to common management practices such as limit feeding and weaning (Budzynska & Weary, 2008; Jasper et al., 2008) has previously been assessed using measures of activity. De Paula Vieira et al. (2008) reported that calves fed restricted amounts of milk spent 1 h longer standing than those fed ad libitum milk, indicating that standing time could be used as a behavioural indicator of hunger. In the present study, there was no measurable difference in lying time throughout the pre-wean period between treatments. This could suggest either that reduction of feeding frequency did not evoke a measurable response or it could be that the individual housing system imposed restricted measures of activity. As such, studies in which the number of focal calves is increased or where calves are group housed could have different results. Studies comparing abrupt versus gradual weaning (e.g. Budzynska & Weary, 2008) have reported an increase in activity in abruptly weaned calves following withdrawal of milk. The decrease in total daily lying time in TAD calves compared with that in OAD28 calves could suggest a behavioural response to weaning in TAD calves and was likely representative of increased time standing to consume concentrate. Concentrate intake was increased in OAD28 calves prior to weaning, and thus, the proportion of time spent standing did not increase at the same rate as TAD calves, potentially indicating that OAD28 calves were better able to adapt to the nutritional change imposed at weaning (Jasper et al., 2008). Calves in the present study were established on solid feed prior to weaning, consuming on average 1.3 kg DM of concentrate on the day preceding withdrawal of MR; this potentially mitigates the behavioural response to weaning.

Heart rate and heart rate variability

Measures of HR and HRV have previously been used to examine physiological response to stressors in dairy cattle (Kovacs et al., 2014). Although HRV can be measured in several ways, RMSSD, a measurement based on the time between each heartbeat, is considered as one of the most informative methods (Clapp et al., 2014). This is because it is linked to vagal tone, which is a useful indicator of how an animal responds to stress (von Borell et al., 2007; Kovacs et al., 2014). Calves under internal or external stress have previously been reported as having an increased HR, decreased RMSSD and therefore lower vagal tone when compared to non-stressed calves (Mohr et al., 2002). It could be hypothesised, therefore, that imposing a reported nutritional stressor (Hulbert et al., 2011) such as reducing milk feeding frequency from twice to once daily would result in a demonstrable physiological response, in this instance, a decrease in RMSSD. Thus, the decrease in RMSSD values of OAD14 calves in Period 1 when compared that of OAD28 calves could potentially be indicative of a stress response to reduction in milk feeding frequency at 14 d of age. Clapp et al. (2014) reported increasing RMSSD values in the 5-d period following weaning, indicating that calves were becoming less stressed; however, in the current study, RMSSD values on the day of and following weaning were consistent to those found during the pre-wean period. This suggests that, in terms of HRV, calves reared according to the management strategies described did not elicit a measurable response to imposed nutritional stressors; being reduction of milk feeding frequency at 28 d of age and weaning. The results observed in OAD14 calves following the reduction in feeding frequency could indicate that imposing once-daily milk feeding at 14 d of age is potentially perceived as a stressor and requires further investigation.