Effects of feeding treatment on growth rates and consequences on performance of primiparous Holstein dairy heifers

The objective of this study was to investigate the effects of feeding rearing programs aiming a first calving between 20 and 27 months (mo) of age on growth, reproduction and production performance of Holstein cows at nulliparous and primiparous stages. Our hypothesis was that, in a seasonal autumn calving strategy, it was possible for late-born heifers in the season to catch up with the growth of heifers born earlier and be inseminated at the same time-period, at a 370 kg body weight (BW) minimum. This would result a first calving age at about 21 to 22 mo, without impairing their later performance. To answer this question, an experiment was run, involving a total of 217 heifers over 3 years. These heifers were split into 3 groups: the first group received a control feeding treatment (SD), the second one an intensive-plane diet (ID1) from birth to 6 mo, and the last group an intensive-plane diet until 1 year of age. Groups SD and ID1 comprised heifers born from September until end of November; ID2 was composed of heifers born later. The present study showed that late-born heifers (ID2) could catch up with the growth of the other thanks to feeding treatment, although there were still 42 kg lighter than both SD and ID1 ones at first calving. There was no difference in reproductive performance of the heifers between the groups. Once primiparous, the cows reared with the ID2 treatment tended to produce less milk than SD and ID1 (about −400 kg over the lactation), and there were no difference regarding milk quality, feed intake, body condition score, or BW. A classification on age at first service was created a posteriori leading to 3 classes with heifers first inseminated at about 12.5 mo (AFS12.5), 14.0 mo (AFS14.0), 15.5 mo (AFS15.5) of age. Heifers in AFS12.5 had a faster growth than those in AFS14.0 and AFS15.5. Once primiparous, the AFS12.5 cows tended to produce less milk at peak than AFS14.0 and AFS15.5 (about −1.5 kg/d) although there was no difference regarding total milk production over the lactation. There was no difference between these groups regarding milk quality, feed intake, body condition score, or BW. All these results support the conclusion that it is possible, through feeding treatment, to help late-born heifers to catch up with the growth of other. This leads to an earlier first calving but do not impair their reproductive and productive performance. Implications Increasing the growth rate of dairy heifers decreased their age at puberty, potentially reducing age at first calving, and ultimately shortening the non-productive rearing period. Heifers first calving at 22.5 months (mo) of age or less presented similar performances than those calving at 23.8 mo of age or more.


48
In seasonal calving systems, heifers usually first calve at a young age (around 24 months(mo)) but . Increasing the nutrient uptake and thus the growth rate for these late-born heifers is a 52 solution to lower this risk. A high growth rate during rearing is associated with a decreased age at 53 puberty and, consequently, 1 st calving may occur as early as 20 to 21 mo of age. Tozer (2000) 54 concluded that a higher plane of nutrition incurred higher daily feed costs, but these costs were 55 recouped when heifers calved at a younger age through savings on labour, housing and overall feed 56 costs. Regardless of the strategy of rearing (group-calving or not), it is, however, necessary for 57 animals to have achieved an adequate body size before calving or milk production potential in the 58 first lactation is compromised (Bach and Ahedo, 2008). Indeed, accelerated growth program for dairy 59 heifers cannot be resumed to puberty attainment. Many authors have studied the impact of growth 60 intensity on further performances (see the literature review of Le Cozler et al., 2008), but if most of 61 them indicated negative impact of too high a growth, some authors indicated limited impact. According 62 to Pirlo et al. (1997), reducing age at first calving to 23 to 24 mo was the most profitable procedure, 63 but not less than 22 mo (except in cases of low milk prices and high rearing costs). They concluded 64 that reluctance to decrease age at first calving is generally attributable to the belief that early calving 65 is detrimental to milk yield and longevity. Here, we designed and led an experiment to determine the 66 effects of feeding treatments on growth parameters, reproduction and production performance of 67 Holstein primiparous heifers first calving between 20 and 27 mo of age, in a seasonal calving system.

68
We hypothesised that genetic improvement over the last decades in dairy production resulted in 69 animals that could calved now at an earlier age than 24 mo of age and results from autumn groups 70 calving strategy could be used in a non-grouping strategy. We focused on the possibility for late-born 71 heifers to catch up with the rest of the heifers at 1 st artificial insemination (AI) at a minimum BW of 72 370 to 380 kg, which resulted in an age at first calving lower than 22 mo. treatments (according to birth order) and fed either a standard diet (SD) or an intensive-plane diet 83 (ID1) from 0 to 6 mo of age. It was expected that, thanks to the feeding intensity chosen, heifers fed 84 SD and ID1 diets would reach 190 to 200 and 220 to 230 kg at 6 mo of age, respectively. Heifers born 85 after 1 st of December (ID2) received the same intensive-plane diet as ID1 heifers from 0 to 6 mo of 86 age, to limit the possible confusion effect between age and treatment during this period.

152
During the entire experiment, all heifers and cows housed indoor had free access to fresh water.

153
Age at first service 154 A classification on age at first service (AFS) was created a posteriori in order to better understand 155 which factors are leading to different AFS and how future performance can be related to AFS.

156
Three classes were made so that the number of animal in each of them is balanced (Table 2).

185
Daily feed intake was calculated individually as the difference between daily feed allowance and 186 refusals. Refusals were collected every day at 7.00 and weighed. The composition of refusal and 187 allowance were presumed to be the same. Dry matter (DM) for silage was determined 5 times per 7 week, while DM of the pellets was determined once a week. Feed composition was estimated based 189 on average samples for corn silage, straw, soybean, and concentrate. No such information was 190 available for fresh grass (see Table 1

Calculations and statistical analysis 226
All data handling and statistical analyses were performed in R using either the lm procedure for 227 ANOVA or the glm for logistic regressions (R Core Team, 2019). Quantitative traits (i.e. growth, ages, 228 live weight, milk production, body condition score, CLA and cycle lengths) were studied using the 229 following ANOVA model :

Growth and reproductive performance of heifers 251
The average BW at birth was 41 kg (±5.2) and was balanced across the groups (i.e. not associated 252 to neither feeding treatment nor AFS). were heavier than these in both the SD and ID1 treatments (229 kg vs 213 kg and 217 kg at 6 mo of 260 age respectively; P < 0.001; Figure 1.A). The highest ADG was found for ID2 heifers from 0 to 6 mo 261 (1046 vs 958 and 976 g/d for ID2, SD and ID1 respectively; P < 0.001). This difference was still 262 significant from 6 to 12 mo of age (789, 703 and 699 g/d for ID2, SD and ID1 heifers, respectively).

263
However, from 12 to 18 mo of age, ADG was significantly reduced for ID2 heifers in comparison of 264 SD and ID1 heifers (660 vs 800 and 774 g/d respectively).

265
Reproductive performance was not affected by the feeding treatment (Table 3), although the number 266 of service tended to be lower for ID2 heifers than for SD and ID1 heifers (1.5 vs 1.9 and 1.8, 267 respectively). Indeed, cows in the 3 feeding treatment showed a similar interval from the start of the 268 breeding season to the 1 st service (13.5 d), similar success at 1 st service (about 62% of pregnant 269 heifers) and similar pregnancy rate by the end of the breeding season (94%).

270
There was no difference in calf BW (37.9 kg) despite a difference in their dam's BW at both 1 st service 271 and 1 st calving (ID2 heifers were lighter than both SD and ID1, Table 3 and Table 5). Heifers in the 272 ID2 treatment calved at a younger age than those in the SD and ID1 treatments (about 2 months 273 earlier, P<0.001; Table 3).

274
Heifers inseminated at a young age (12.5 mo of age on average; AFS12.5) tended to have a higher 275 growth rate than heifers inseminated at either 14.0 (AFS14.0) or 15.5 (AFS15.5) mo of age, from 0 to 6 276 mo of age (1001 vs 960 and 978 g/d; P 0<0.10; Table 4). This difference became more important 277 from 6 to 12 mo of age (759 vs 688 and 698 for AFS12.5, AFS14.0 and AFS15.5 respectively; P < 0.01    Figure 1.B). Growth was reduced for AFS12.5 animals from 12 to 18 mo of age, with an ADG 280 of 712 g/d, compared to 799 and 790 g/d for AFS14.0 and AFS15.5 (P < 0.001; Table 4). This is 281 consistent with the effects of feeding treatment observed, and with the distribution of animals among 282 the classes of AFS and feeding treatments (Table 2). 283 Fertility was not affected by age at first service (Table 4)

286
There was no difference in calf BW (37.9 kg) despite a difference in their dam's BW at 1 st service and 287 at 1 st calving (AFS12.5 heifers were lighter than the ones in AFS14.0 themselves lighter than the one in 288 AFS15.5, Table 4 and Table 6). Consistent with the age at 1 st service, heifers in the AFS12.5 group 289 calved younger than those in the AFS14.0 who calved at a younger age than those in the AFS15.5 group 290 (  Figure 2.A.), which is consistent with the fact that ID2 heifers 295 first calved at a younger age than SD and ID1 heifers (Table 4). No difference between the feeding 296 treatments was noticed in BCS during the first lactation. On a 308 d basis, ID2 cows tended to produce 297 less milk than SD and ID1 cows (6920 vs 7312 and 7370 kg; P < 0.10; Table 5; Figure 2.C.). There 298 was no difference between feeding treatments regarding average fat and protein contents. Milk yield 299 peak was reduced for ID2 cows compared to both SD and ID1 ones (28.7 vs 31.3 and 31.9 kg; P < 300 0.001). During the first 7 weeks of lactation, ID2 cows were lighter (on average -38 kg compared to 301 SD and -25 kg compared to ID1); and produced less milk (-3.1 kg/d compared to both SD and ID1).

302
This difference was already shrunk during the last part of the period (from 8 to 15 weeks), ID2 cows 303 weighed 27 kg and 17 kg less than SD and HD1 cows respectively; and produced 2.2 kg/d and 2.9 304 kg/d of milk less than SD and HD1 cows respectively.

305
The feeding treatment of dairy cows during the rearing period did not affect ovarian cyclicity during 306 the 1 st lactation (Table 5). On average, the CLA was about 20.4 d, the first IOI was about 20.7 d with 307 no difference among treatments concerning the LUT and the ILI. There was no difference concerning 308 the subsequent cycles neither, with an IOI of 23.3 d on average. The distribution of abnormal pattern 309 of ovarian activity was not significant, although the ID2 cows showed a lower proportion of normal 310 profile than ID1 cows, that had themselves a lower proportion of normal profiles than SD cows (53% 311 vs 59% vs 65% respectively; Table 5). ID2 cows had an incidence of 33% of PLP abnormalities, while 312 the incidence in ID1 and SD cows were 18 and 19% respectively (Table 5). About 86% of the cows 313 were pregnant at the end of the breeding season, with no relationship with feeding treatment. Although 314 the difference in cyclicity between feeding treatment did not impair the re-calving rate, ID1 cows 12 needed more inseminations to be pregnant than SD cows (2.4 vs 1.9; P < 0.05; Table 5). The number 316 of services needed to achieve pregnancy was about 2.2 for the ID2 cows. Subsequent calf BW was 317 not affected by the feeding treatment.  Table 6). 331 332

333
Fertility of cows was very little affected by AFS. Concerning ovarian cyclicity, all 3 groups of AFS 334 showed a similar CLA, with similar cycles length, except for cows in AFS15.5 that tended to show longer 335 ILI between the 2 nd to 4 th cycle, than cows in AFS12.5 and AFS14.0 (Table 6) Cows in the AFS14.0 group 336 showed a lower incidence of PLP than cows in AFS12.5 and AFS15.5 (8% vs 29% and 23% respectively; 337 P < 0.05; Table 6). Fertility was not affected by AFS neither: all groups showed a similar number of 338 inseminations (2.2 on average) and 86% of the cows were pregnant at the end of the breeding season.

341
Feed intake was not different neither between feeding treatment, nor between AFS groups (17 kg 342 DM/d; Appendix 1). Morphological traits analysis based on age at first calving cohorts 2009-10 and  BW of animals at 6 mo of age were high (111 cm heart girth and 220 kg body weight, for example), 375 which fits well with recommendation for optimal age at first calving at 24 mo of age or less. In a study of Ettema and Santos (2004) (Bartlett et al., 2006). Increased nutrient intake is also associated with increased plasma IGF1 (Smith  life, but the differences in milk yield were most apparent during early and peak lactation. In particular, 399 higher calf growth rates were not significantly associated with future milk yield but were associated 400 with higher BW in lactating cows and higher odds of survival to first lactation. When milk lactation was 401 corrected to BW, no difference was found in milk yield or composition, regardless of rearing treatment.

403
Decreasing age at first calving is an effective way to decrease the length of non-productive days 404 during rearing and first calving at around 24 months of age appears to be is optimal for profitable kg less milk on a 305 d lactation length basis. In present study, we also noticed that young heifers 409 produced less milk during the first part of lactation, but the total milk yield over 305 d was not different.

410
However, it could be noticed that despite no difference from the statistical point of view, the difference 411 was very similar when age at first calving decreased from 24.8 to 23.8 mo of age: 134 kg less on a 412 305 d basis.
In present study, fertility was not affected by age at first service. In a previous paper on puberty 415 attainment based on 2011-12 cohort, we noticed that most heifers reached puberty before oestrus 416 synchronisation, at an average age of 10.3 ± 2.2 mo (6.2 to 14.4 mo), averaging 296 ± 40 kg (224 to 417 369 kg) BW (Abeni et al., 2019). It occurred one mo earlier for ID2 heifers than in SD and ID1 heifers.

418
Puberty onset at 9 to 10 mo of age or less meant that 3 or 4 oestrous cycles occurred before 419 insemination, which is generally consistent with good fertility results in many species (Lin et al., 1986; 420 Byerley et al., 1987;Robinson, 1990;Le Cozler et al., 1999). Regardless of calving strategy, lowering 421 the age at puberty and, consequently, the age at first insemination means that it is an efficient way to 422 shorten the non-productive period before calving. However, as suggested by Meyer et al. (2006), it 423 might reduce pre-pubertal mammary gland development by shortening the allometric phase of 424 mammary gland growth and, in some cases, impair further milk production. Similar to fertility in heifers, 425 fertility in primiparous cows during first lactation was not affected with age at first calving. Wathes et 426 al. (2008) reported that optimal fertility and maintenance of maximum performance in the first lactation 427 were reached at the calving age of 24 to 25 mo, although heifers that calved at the age of 22 to 23 428 mo were the best in overall performance and longevity over 5 year, partly because heifers with good 429 fertility also had a high level of fertility as cows.

431
Finally, In the present paper, we also noticed that for a similar feed allowance, early calving heifers 432 ate a similar amount of feed, produced less milk and at the end, were able to catch the difference in 433 BW and development. All these results indicate that, as already reported by Krpálková et al. (2014), 434 the objective of a rearing period leading to an age at first calving less than 23 mo of age in Holstein 435 heifers proves to be a suitable option for successful rearing of heifers with optimal subsequent 436 production and reproduction in a herd with suitable management.

Declaration of interest 443
The authors declare that the research was conducted in the absence of any commercial or financial 444 relationships that could be construed as a potential conflict of interest.

Ethics statement 447
Experimental work has been conducted in accordance with French national legislation on the use of 448 animals for research. Protocol received agreement (00944-02) from French Ethical Committee n°7.

Software and data repository resources 451
None of the data were deposited in an official repository.