• Keine Ergebnisse gefunden

The application of behaviour sensors and sward measurement to support grazing management

Im Dokument roles of grassland in the European (Seite 111-114)

Timmer B.1, Zom R.L.G. 2, Holshof G.2, Spithoven M.1 and Van Reenen C. G.2

1Wageningen University, Animal Nutrition Group, De Elst 1, 6708 WD Wageningen, the Netherlands;

2Wageningen UR Livestock Research, Animal Nutrition, De Elst 1, 6708 WD Wageningen, the Netherlands; ronald.zom@wur.nl

Abstract

The aim of this study was to evaluate whether commercially available accelerometers for the recording of grazing time and cow activity (movements, lying, steps) combined with data on pre-grazing sward surface height, and cow characteristics (parity, days in milk, milk yield, milk composition) could be used to estimate intake of herbage by grazing dairy cows. Sixty cows which were involved in a grazing trial were equipped with two 3-dimensional accelerometers. During one test week individual herbage and total dry matter intake were measured using the n-alkane technique. The collected data were, pre-grazing sward height (cm), time spend grazing, standing and lying in seconds per quarter and number of steps per quarter. Regression analyses indicated that besides accelerometer data, inclusion of cow characteristics improved the accuracy the prediction of herbage dry matter intake. These preliminary results indicate that the use commercial available three-dimensional accelerometers may have a potential as a tool to monitor herbage intake at grazing Keywords: cow behaviour, herbage intake, grazing management

Introduction

Three-dimensional (3D) accelerometer sensors can be used to monitor cow behaviour and differentiate between the time that cows spent grazing, ruminating, lying, walking, standing, number of movements, acceleration etc. The behavioural data derived from 3D-accelerometer sensors have shown to be very valuable as a supportive tool on dairy farms for heat detection and monitoring health problems related to claws and legs. However, it is an intriguing question whether 3D-accelerometer sensor data can also be applied to estimate herbage dry matter intake (HDMI) of individual cows at grazing. A good estimate of grazing HDMI of individual dairy cows is necessary to balance supplemental feeding (concentrates and forages) and herbage allowance (kg DM herbage cow-1). Moreover, monitoring of the grazing HDMI by the herd enables assessment of herbage production and utilization. The aim of this study was to investigate if individual HDMI can be estimated from 3D-accelerometer sensor data (grazing time, standing, walking) and if the estimation of HDMI could be improved by combining grazing time with sward height measurements, and animal performance data.

Materials and methods

Sixty dairy cows (21 primiparous and 39 multiparous) were divided in 20 blocks of three dairy cows according to similarity in lactation number, days in milk, fat and protein corrected milk production and body weight. The cows of each block were randomly assigned to one of three grazing systems. Grazing systems were strip grazing (SG), ‘rotational’ continuous grazing system (RCG), rotational grazing on 24 1-day paddocks (DRG) which are described in detail by Holshof et al. (2016; this book). Cows were milked twice daily at 0600 and 1600 h using a mobile milking parlour and individual milk yield was recorded throughout the experiment. Individual milk samples were collected weekly at two consecutive milkings and analysed for fat, protein, lactose and urea (Qlip laboratories, Zutphen, the Netherlands).

All cows were individually supplemented with 4.6 kg cow-1 d-1 of a standard commercial concentrate fed in two equal portions during milking. The cows of treatment groups of cows SG and RCG were

fed maize silage at rate of 6 kg DM maize silage cow-1d-1. Daily, per group, DM content and the fresh weight of maize silage were recorded. Daily, pre-grazing herbage allowance (PrHA; kg DM ha-1) of each treatment was estimated from the sward surface height (SSH) measured by a rising-plate meter.

The regression of PrHA against SSH was derived from 76 plots cut and the following relationship was obtained: PrHA (kg DM ha-1) = 314×SSH (cm) – 1120; r2 =0.93). All cows were equipped with two commercially available 3D accelerometers. One device, the ‘IceQube’ (IceRobotics, UK), was attached to a hind leg and recorded total standing time, lying time and number of steps. The second device, the Smarttag Neck (NEDAP, Groenlo, Netherlands) was attached to the neck and recorded grazing time in sec h-1 at pasture. Individual HDMI and maize silage dry matter intake (MDMI) was determined using the n-alkane method of Dove and Mayes (1991). During a 14 day dosing period, the cows received twice daily with 0.4 kg of a concentrate containing 1,046 mg kg C32 n-alkane, at each milking. During the same period the maize silage was labelled with soybean meal which was enriched with 4,165 mg C36 n-alkane, providing 161 mg C36 n-alkane kg DM-1 maize silage.

From day 7 to 14 of the alkane dosing period the herbage, maize silage and concentrates were sampled daily and pooled by grazing treatment for the whole sampling period. During day 7 to 14 of the dosing period, the faecal samples were, if possible, collected opportunistically from each cow twice daily after each milking. When cows were not observed defecating, then faeces were collected by rectal stimulation. The faeces samples were pooled to one sample for each cow. The concentrations of n-alkanes in grazed grass, maize silage, concentrates and faeces was analysed according to the procedures described by Abrahamse et al. (2008). The HDMI was calculated according to Dove and Mayes (1991) on basis of the concentrations C32 and C33-alkanes in faeces, herbage, maize silage and concentrates. The MDMI was calculated on the basis of the concentrations of C35 and C36 n-alkanes in the faeces, herbage, maize silage and concentrates.

A general linear modelling procedure (GLM) was used to predict HDMI, which was considered as the response variable (Y), with cow behaviour data (standing time, lying time, number of steps, grazing time), milk yield and milk composition, milk urea pasture data (SSH, PrHA grazing area) as explanatory variables. Stepwise multiple regression analysis was used to identify the best fitted models, based on the percentage of adjusted variance explained (adj. r2) and best precision (lowest Mallows CP).

Results and discussion

The best fitted model for HDMI is presented in Table 1. Cow behaviour, expressed as eating time per hour at pasture, grazing area per cow, days in milk, milk yield, protein concentration and milk urea content were included in the final model for the prediction of HDMI. Standing time, lying time and number of steps did not contribute to prediction of HDMI. Inclusion of SSH did not improve the model, since the grazing area per cow and SSH were highly correlated. Inclusion of the grazing area per cow in the model was more convenient since the size of the grazed strips or paddocks already known by the farmer, and SSH required additional measurements. The proposed model resulted in a close correlation between herbage intake measured with the n-alkane technique and herbage intake predicted using behaviour sensors, sward and animal performance data (Figure 1). Inclusion of animal parameters such as days in milk and milk yield may be related to the cows’ physiological status and energy requirements. The effects of milk protein content and milk urea on predicted HDMI could be related to the supply of dietary protein to the cow.

Rumen degradable protein is the main source of nitrogen for microbial protein synthesis and finally the main source of amino acids for milk protein synthesis. A surplus of rumen degradable protein is broken down to ammonia which passes through the rumen wall and is transported through the portal vein to the liver where it is converted to urea. Urea is partly recycled to the rumen with saliva and partly excreted in milk and urine. In the current experiment, grazed grass provided the largest proportion of the total rumen

degradable protein intake by the cows. Therefore, the explanatory effect of milk urea content is associated with higher intake of rumen degradable protein. This also implies that the model should be used with caution when grazed grass is supplemented with feeds high in rumen degradable protein.

Conclusions

The study shows that HDMI of individual cows measured with the n-alkane can be reliably estimated from eating time, grazed area, milk yield, milk protein and milk urea. Using this information may have potential as management tools to monitor HDMI and grassland utilization. However, further research is needed to validate the proposed model with independent data.

References

Abrahamse, P. A., J. Dijkstra, B. Vlaeminck, and S. Tamminga (2008) Frequent allocation of rotationally grazed dairy cows changes grazing behavior and improves productivity. Journal of Dairy Science 91, 2033-2045.

Dove, H., and R. W. Mayes (1991) The Use of Plant Wax Alkanes as Marker Substances in Studies of the Nutrition of Herbivores – a Review. Australian Journal of Agricultural Research 42, 913-952.

Holshof, G., Galama, P.G. and R.L.G. Zom (2016) Development of three grazing systems combined with high stocking rate; 2015 results. (This book).

Table 1. Model parameter estimates with standard error (s.e.) and t probability values of the generalized linear model for prediction of herbage dry matter intake.

Parameter Unit Estimate s.e. t (d.f.) t pr.

Constant µ -23.72 3.68 -6.44 <0.001

Eating time sec h-1 at pasture 0.003288 0.000741 4.44 <0.001

Area per cow ha cow-1 d-1 -162.6 21.2 -7.68 <0.001

Days in milk 0.01433 0.00444 3.23 0.002

Milk yield kg 0.4677 0.0438 10.67 <0.001

Milk protein % 3.79 1.12 3.38 0.001

Milk urea mg 100 g-1 milk 0.2388 0.0522 4.58 <0.001

21 Measured herbage intake {kg DM d-1)

Figure 1. Correlation between measured herbage intake (n-alkanes) and predicted intake using cow and sensor data.

Effects of grazing previously abandoned grassland on

Im Dokument roles of grassland in the European (Seite 111-114)

Outline

ÄHNLICHE DOKUMENTE