Brocard V.1, Danilo S.2 and Allain C.1
1Institut de l’Elevage, Le Rheu, France; 2Institut de l’Elevage, Angers, France; valerie.brocard@idele.fr
Abstract
The Lifecorder+®, a uniaxial neck mounted activitymeter which is used to record grazing time of dairy cows, was validated recently (Allain et al., 2014; Delagarde and Lamberton, 2015). Experiments during spring 2015 on two French experimental automatic milking systems (AMS) farms show a great individual and temporal variability. In a system with feed supplementation (Derval), daily grazing time decreases together with the decline of grass quantity and quality and the increase in the amount of supplementation. Grazing time is about 43% of the access time to pasture. In a 100% grazing system (Trévarez), where daily pasture allowance is constant, daily grazing time is about nine hours. It is longer when grass quality improves. The cows’ hierarchy of access to the AMS is a predominant feeding behaviour factor. ‘Dominant’ animals have longer grazing time and higher animal performances (milk production, milking frequency) compared to ‘dominated’ animals. The use of grazing time data could be of interest in addition to other decision support tools already used by farmers (pasture probe, grazing calendars) to better combine AMS and grazing.
Keywords: grazing time, accelerometer, Lifecorder +
Introduction
The increasing number of automatic milking systems (AMS) in France and in Europe has resulted in a reduction of grazed grass in the diet of dairy cows. To improve the competitiveness of the European dairy farming by combining grazing and AMS, the European project Autograssmilk aimed to enhance the use of new technologies to succeed in this association. Recently, Allain et al. (2014) and Delagarde and Lamberton (2015) showed that a human activitymeter named Lifecorder+® (LC+, Kenz) could be used to assess cows grazing time. This technology was used to record daily grazing time of dairy cows in two systems combining AMS and grazing, one with supplementation and the other one in full grazing.
Materials and methods
The LC+, is a device for monitoring uniaxial acceleration. To assess grazing time on dairy cows, the sensors were mounted during two months on neck collars on cows from two AMS experimental farms in Western France (23 out of 67 cows equipped in Derval farm – Pays de Loire, 14 out of 46 cows equipped in Trévarez farm – Brittany). In Derval, the grazing system is simplified (three paddocks of 8, 10 and 10 ha) with buffer feed supplementation. In Trévarez, a rotational 100% grazing system with three paddocks per day is applied. The grazing organisations on the two farms are presented in Figure 1. The cows equipped with the LC+ were chosen to be representative of the herd in terms of production level, parity and lactation stage. The data from the sensors were converted into grazing time when the activity level was exceeding a certain threshold (Allain et al., 2014). In Derval, an electronic identification system at the entrance and the exit of the barn was used to calculate the time spent by cows in the pastures. In Trévarez, the AMS was in the pastures (mobile robot), so the cows were always outside except when they were milked. Finally, 1,323 recordings were available for Derval and 518 for Trévarez. A typology of animals according to their grazing behavior was achieved through the SPAD 8.2.10® software (groups A to D in Derval and A to C in Trévarez). A principal components analysis (PCA) was used to calculate similarities between individuals. Hierarchical ascending classifications (HAC) followed by descriptions of the classes were achieved to make groups of individuals with common features. Animal characteristics
of the different classes of cows were tested by an ANOVA (proc GLM) with the SAS 9.4® software.
The values presented (Table 1) are adjusted averages. The differences between groups were evaluated by Tuckey and Fisher tests.
Results and discussion
In a full grazing system (Trévarez), cows spent 533 minutes grazing per day in average (Table 1). In a system where they are supplemented with maize silage (Derval), the pasture access is limited and they spent only 319 minutes grazing. The difference between the two systems was also observed for the number of meals (number of grazing sequences per day – 6.8 vs 5.3) and the meals duration (grazing time per pasture visit – 84 vs 64 min). The ratio between the grazing time and the access time to pasture was quite similar in the two systems (37 vs 43%) and close from the one observed by Kaufman et al. (2009).
In both farms, the results showed a great individual and temporal variability among the data. In Derval, grazing time decreased with the decline of grass quantity and quality and an increase in the amount of maize silage delivered. In Trévarez, where pasture allowance remains at the same level every day, grazing time is longer when grass quality is better.
Results of the PCA and the HAC are also presented in Table 1 with the comparison between the different groups of animals. In both farms, we observed that feeding behaviours enabled to bring out the place of animals in the herd social order. ‘Dominant’ animals, which have the first access to the AMS and therefore to the pastures, have longer grazing time and higher performances (groups A and C in Derval and group A in Trévarez) compared to ‘dominated’ animals (groups D in Derval and B in Trévarez). The Table 1. Descriptive statistics and groups’ comparison in Derval and Trévarez farms. Groups A to D in Derval and A to C in Trévarez were constituted from a principal components analysis used to calculate similarities between individuals and by a hierarchical ascending classifications followed by descriptions of the classes to make groups of individuals with common features.
Derval farm (n=1,323)
Groups A B C D Herd average (SD) P (Fisher)
Number of animals 6 4 8 5 23
Grazing time (min) 360a 365a 294ab 274b 320 (102) <0.001
Number of meals 5.6a 5.7a 5.2b 4.8b 5.3 (1.8) 0.0064
Meals duration (min) 69a 69a 60b 59b 64 (23) 0.0003
Access time (min) 817 a 776 a 801 a 686 b 775 (237) 0,001
Grazing time/access time (%) 44a 47a 37b 40b 43 (14) <0.001
Exit time 0h45a 2h00b 0h50a 3h00b 1h20 <0.001
Yield (kg cow-1 day-1) 30.0a 21.2b 29.4ab 24.8ab 27.1 0.047
Milking frequency (milk day-1 cow-1) 2.03 1.84 1.96 1.77 1.92 0.16
Trévarez farm (n=518)
Groups A B C Herd average (SD) P (Fisher)
Number of animals 4 5 5 14
Grazing time (min) 565a 525b 508b 530 (79) <0.001
Number of meals 7.2 6.9 6.3 6.8 (1.8) 0.16
Meals duration (min) 82 81 85 84 (22) 0.64
Grazing time/access time (%) 39a 36b 35b 37 (6) 0.0004
Access time 1st paddock 8ha 9h45b 8h20a 8h45 0.018
Yield (kg cow-1 day-1) 21.1a 13.9b 18.9a 17.7 0.045
Milking frequency (milk day-1 cow-1) 2.01a 1.57b 1.77ab 1.77 0.023
other groups (B and D in Derval and C in Trévarez) are intermediary. Figure 1 shows the grazing kinetics of the different groups. In Derval, we can see that the ‘dominant’ animals have a significant meal at night and a big one when the sun is rising. The ‘dominated’ cows are penalized by their late access to pasture and are not able to compensate at sunrise and in the afternoon. In Trévarez, the graph in Figure 1 shows that the ‘dominated’ animals have a late access to the first paddock and often do not go to the 3rd one.
Consequently they have to graze the left-overs of the 2nd paddock.
Conclusions
The use of grazing time data could be of interest in addition to other decision support tools already used by farmers to better combine AMS and grazing. In order to maximize grass ingestion, several recommendations can be provided. First of all, forage supplementation must be calibrated to the grazing allowance and not the opposite. Then, the access to AMS, therefore to pasture, could be regulated according to the animals’ social hierarchy in order to give an advantage to ‘dominated’ cows.
References
Allain C., Raynal J., Beck C., Delagarde R. and Brocard V. (2014) Use of the Lifecorder +® sensor to assess grazing time of dairy cows.
Grassland Science in Europe 20, 166-168.
Delagarde R. and Lamberton P. (2015) Daily grazing time of dairy cows is recorded accurately using the Lifecorder Plus device.
Applied Animal Behaviour Science 165, 25-32.
Kaufman L.D., Muenger A., Rerat M., Junghans P., Goers S., Metges C.C. and Dohme F. (2009) Dépense énergétique, activité physique et comportement alimentaire des vaches laitières au pâturage ou consommant la même herbe à la crèche. Rencontres Recherche Ruminants 16, 33-36.
Figure 1. Grazing organisation and grazing behaviours of the cow groups in Derval (left) and Trévarez (right).