Estimating performances of dairy buffaloes in a new model of automated milking system

Abstract

Background: In recent years automatic milking systems (AMS) have emerged as a technology that can improve traditional milking routine. AMS offer several benefits, such as regular milking routine, health of the mammary glands and labour efficiency, higher milk yield (MY), and quality. In contrast with dairy cows, the application of AMS in dairy buffalo herds remains underexplored. This study aimed to extend the current knowledge regarding AMS and evaluate for the first time the efficiency of new AMS models in buffalo species on a wide dataset, by focusing on MY and quality. Two sources of data were analysed: data collected at each milking by the AMS software for a duration of 22 months, and monthly milk yields and quality traits. The statistical analysis was performed through the R software. A mixed model for repeated measures was adopted, using days in milk (DIM) and parity as fixed factors and the visit/day of sampling as repeated measure. A linear regression model was also used to study the relationship between the number of milking per buffalo per day (as independent variable), lactation persistency, DIM, parity, somatic cell score (SCS), and MY (as dependent variables).

Results: The effects of parity, DIM and their interaction were significant for all milk quality traits (monthly recordings), except for their interaction on somatic cell count (SCC) and SCS. An average of 2.35 milkings/buffalo was recorded, with a mean milking duration of 7.36 min. The average MY was 9.15 kg/day, with 7.97% and 4.81% fat and protein content, respectively and the lactation peak occurred at 38.17 ± 1.31 DIM. The difference between multiparous and primiparous was clear in terms of both MY and energy-corrected milk (ECM). Also, parity and DIM significantly influenced both average milk flow rate (1.49 kg/min on average) and peak milk flow rate (2.68 kg/min on average): these showed a comparable trend describing higher values during the first days in milk followed by a decrease throughout the lactation, along with a decreased milk production.

Conclusions: The present study was the first to investigate new model of AMS in buffalo species in a large dataset providing insight into the new model of AMS as an effective milking system for buffalo species. Still, further research is encouraged to confirm these findings, as well as to compare the performances of different automatic milking systems, possibly under different management practices and environmental conditions.

Keywords: Buffalo husbandry; Italian mediterranean buffalo; Milk quality; Milking performance; Precision livestock farming.

Fig. 1

Trend for the number of milkings (arithmetic means, no./buffalo per day) across classes of days in milk (DIM) and parity (1–5+) in the Italian Mediterranean Buffaloes (n = 244,536). Class A from 1 to 30; class B from 31 to 60; class C from 61 to 90; class D from 91 to 120; class E from 121 to 150; class F from 151 to 180; class G from 181 to 210; class H from 211 to 240; class I from 241 to 270; class L from 271 to 300 DIM

Fig. 2

Trend for milk flow peaks (panel A) and milk flow rate (panel B) (arithmetic means, (kg/min)) across classes of days in milk (DIM) and parity (1– 5+) in the Italian Mediterranean Buffaloes. Class A from 1 to 30; class B from 31 to 60; class C from 61 to 90; class D from 91 to 120; class E from 121 to 150; class F from 151 to 180; class G from 181 to 210; class H from 211 to 240; class I from 241 to 270; class L from 271 to 300 DIM

Fig. 3

Trend for milk yield (kg/day, panel A), fat and protein content (%, panels B and C, respectively), energy corrected milk (arithmetic means, panel D) and SCS (panel E) across classes of days in milk (DIM) and parity (1–5+) in the Italian Mediterranean Buffalo. Class A from 1 to 30; class B from 31 to 60; class C from 61 to 90; class D from 91 to 120; class E from 121 to 150; class F from 151 to 180; class G from 181 to 210; class H from 211 to 240; class I from 241 to 270; class L from 271 to 300 DIM