Top Navigation Bar

ASAE Conference Proceeding

This is not a peer-reviewed article.

Model for Predicting Milk Production in Jersey Cows in Hot Weather

L, A. Laloni, I. A. Nääs, M. Macari, D.F. Pereira and M.G. Pinheiro

Pp. 320-324 in Fifth International Dairy Housing Proceedings of the 29-31 January 2003 Conference, (Fort Worth, Texas, USA), ed. K. A. Janni. ,Pub. date 29 January 2003 . ASAE Pub #701P0203

Abstract

In Brazil the adoption of semi-confinement leads to special adapted management methods in dairy production, and milk production can be improved by the use of technology that assures better herd management. Index variables such as air temperature, relative humidity and air speed were considered. The values of rain index, solar radiation and pasture soil temperature were also considered as stress agents that could potentially influence milk production. The objective of this research was to identify the variables that potentially would interfere in the milk yield in Jersey cows, and to develop an index for predicting milk production for high productivity milking cows lodged in pasture under hot weather conditions. The experiment took place in Ribeirão Preto, in Southeastern Brazil, latitude 210 11’ S, longitude 470 48’ W, and altitude 621 m, using two treatments with 14 Jersey cows each: Treatment A – cows waited for 30 minutes prior to milking in a room with a shower and a fan, and Treatment B – the cows did not have access to this room (control). At other times the cows had access to a pasture. Differences in the effect of the treatment in the average milk yield were not statistically significant for the treatments. The analysis for studying the effect of designing the model led to a probabilistic model relating the variables milk production and rain index coeficient: Milk Production (kg) = 10.67 (kg) + 0.017 (kg/mm).

KEYWORDS. Environment, dairy cows, model.

Introduction

When the thermal balance is inadequate, it affects an animal’s health and productivity. Consequently, the thermal environment is extremely important to animal production. Each animal species has an optimal thermal environment in which maximum productivity occurs. This ideal thermal environment is referred to as the animal’s thermoneutral zone.

Thermal Comfort Indexes can be used to characterize or quantify different thermoneutral zones appropriate to different animal species. Thermal Comfort Indexes represent, in a single mathematical model, factors that characterize both the surrounding thermal environment and the stress caused to the animals by the environment. They are extremely useful to producers because they offer a method of calculating the thermal stress response of their animals with a single mathematical model using as input the climatic conditions at a given time. The Temperature Humidity Index developed by Thom (1958) for humans is often used as a measure of thermal comfort for livestock. Berry et al. (1964) and Buffington et al. (1981) have developed indexes specifically for lactating dairy cows.

The Equivalent Temperature Index (ETI), developed by Baêta (1987) was based on heat loss rate and on milk production. The ETI was based on data collected in a climatic chamber at the Brody Animal Climatology Laboratory, University of Missouri, Columbia. Its limits for use were established as dry bulb temperatures between 16 and 41 ° C, relative humidity between 40 and 90%, and air velocities between 0.5 and 6.5 m/s. ETI was defined in Baêta (1987) as:

ETI = 27.88 – 0.456 (t) + 0.010754 (t)2 - 0.4905 (RH)+ 0.00088 (RH)2 + 1.1507(v) - 0.126447 (v)2 + 0.019876 (t)(RH) - 0.046313 (t)(v) , where:ETI = Equivalent temperature index; t = air temperature( ° C), RH= air relative humidity (%) , and v = air velocity (m/s).

The computer software MILK PLUS, developed by Nääs (1991), used the concept cited in Baêta (1987) as a starting point. MILK PLUS evaluates the heat balance within freestall housing and predicts milk production for the tropical climatic conditions found in Brazil. Approximately 90% of the dairy cattle in Brazil are housed in freestall barns that allow access to pasture. This housing system is referred to as partial freestall housing and is intensively used mainly during the dry season. The building characteristics are of open sided, average height of 4m, freestall area of 2 m2 per cow, with stall dimensions of 2m length by 1m of width, using one stall per cow. MILK PLUS was tested on Holstein cattle in Southeastern Brazil (Nääs et al., 1991). The observed milk production was not statistically different from the predictions of MILK PLUS when cows were totally confined in freestall housing, mainly during the dry season (Nääs et al., 1994). However, when cows were allowed to graze during the day, milk production differed from the predictions of MILK PLUS.

When cows stay on pasture most of the Thermal Comfort Indexes do not predict milk yield well. Brazilian milking cows stay in the so-called semi confinement barn, on pasture or under shades, and in a covered holding area, prior to milking in the milking parlor. They stay around 30-45 minutes in this holding area and the environment of this place may affect their milk yield.

Objectives

The objectives of this research were: 1) to evaluate the use of showers prior to milking, 2) identify variables that potentially interfere with the milk yield from Jersey cows, and 3) to develop a model for predicting milk production for highly productive milk cows lodged on pasture under hot weather conditions.

Methodology

The experiment took place in Ribeirão Preto in Southeastern Brazil, latitude 210 11’ S and longitude 470 48’ W, and altitude of 621 m. This region has an average yearly dry bulb temperature of 22 °C and a rain index of 1.416 mm occurring during the Summer. During the hottest season the maximum average dry bulb temperature is 31°C while the minimum average dry bulb temperature is of 19°C. Two treatments with 14 Jersey cows were used: Treatment A – cows waited for 30 minutes prior to milking in a room with a shower and a fan, and Treatment B – the cows did not have access to this room (control). At other times the cows had access to pasture. The cows were milked twice a day and their time in the holding area with the treatments was approximately 30-45 minutes. After milking they stayed in a tree shaded pasture with Tanzania grass. A supplementary ration of corn, soybean and chopped sorghum was provided in mobile feeders.

Environmental variables including dry bulb, wet bulb, and black globe temperatures were collected inside the waiting area as well as in the pasture. Soil temperature and rain index was also measured at the following hours: 11:00h, 13:00h, 15:00h e 17:00h using digital thermometer Instrutherm ® +/- 0,1°C.

Cow rectal temperatures were measured after milking using a clinical thermometer. Wind speed was measured using an anemometer Kestrel ® , 0-30m/s scale and +/- 0,1m/s. Skin temperature was recorded after milking using an infrared thermometer Raytec ® .

Data were analyzed statistically to determine the significance of the variables influencing milking yield, using the software Statistica 5.1 e Stata 5.0, e Spad 3.5. For analyzing the environmental variables the approach used by Johnson and Wichern (1998) was used (the analysis of main components). The interpolation of Minimum Weight Square was applied using the software Statistica 5.0.

Results and discussion

The statistical analysis showed that there was no significant difference in both treatments even though the average milk production in treatment A was 0.41 kg higher than the results in treatment B. This increase in the average values happened during February (0.9 kg) a month with high rain index. The effect of the data variability can be seen in Table 1.

Table1. Average milk production in each treatment during the months of the experiment (n.s. at a=0.05)

Month/ Treat.

Average Production (kg)

Min. Production (kg)

Max Production (kg)

Jan/ A

10.97 ± 4.12

2.8

22.1

Jan/ B

10.75 ± 4.31

4.2

21.6

Feb/ A

11.86 ± 3.97

4.5

19.8

Feb/ B

11.05 ± 4.59

3.6

21.8

Apr/ A

11.13 ± 3.11

3.8

16.3

Apr/ B

11.58 ± 4.75

3.2

18.7

The results of the environmental data were grouped and selected according to their significance and the average and statistical data are shown in Table 2.

Table 2. Average values of selected environmental data.

Variable

Average

St. Deviation.

Min

Max

radinc (W/m2)

386.39

75.56

121.7

504.3

radir (W/m2);

52.34

16.65

3.01

80.6

tamax (ºC)

32.03

3.09

20.01

36.95

tamin (ºC)

18.58

2.05

8.23

21.01

tslmax (ºC)

31.29

3.33

24.09

35.81

tslmin (ºC)

24.91

2.13

16.79

28.01

precip(mm/day)

7.02

11.68

0

47

Where: radinc = Solar Radiation Outside; RADir= Solar Radiation Inside; TAMax= Max dry bulb temperature inside; TAMin= Min dry bulb temperature inside; TSLMax= Max Soil temperature in pasture; TSLMin= Min Soil temperature in pasture; PRECIP= Rain Index

A regression model was used for adjusting milk production taking into account the randomized effects and the most significant variables studied. As the rain index was found to be the most significant effect, as shown in Table 3, a model was described such as:

Milk yield = 10.67 + 0.017 (Rain Index) , where milk yield is expressed in kg, the term 10.67 is the average milk production in kg, the constant 0.017 is expressed in kg/mm and the rain Index is expressed in mm of rain/day.

Confronting the simulated values using the suggested index and the real values of data selected randomly it was found that the equation fits the values with a R2 value of 0.9324, as seen in Figure 1.

Table 3. Simple regression data and random effect in milk production as function of rain index.

Milk yield (kg)

Coef.

St Dev.

z

P > | z |

a =0.05

Rain index (mm/day)

0.017068

0.0088945

1.919

0.055

- 0.0003646

0.0345011

constant

10.67345

0.7561083

14.116

0.000

9.1915

12.15539

Where: z = statistical z test, P= Pvalue, and a value=statistical significance

Figure 1. Milk yield index equation approach

320-324_files/image1.gif

Conclusion

The results indicate that the use of showers prior to milking did not have a significant effect on milk yield, while the rain index effect was relevant. This can be validated when data was checked with the milk yield in February when the rain index and dry bulb temperatures were high. During this month it was found an increase of 0.9 kg in the treatment A indicating that the use of showers may influence milk yield when associated with the cooling effect of the rain in pasture. This model can be used for predicting milk yield under the studied conditions.

Acknowledgements

The authors wish to thank FAPESP for the financial support and the Instituto de Zootecnia for the use of the animals and freestall.

REFERENCES

Baêta, F. C., N. F. Meador, M. D. Shanklin and H. D. Johnson. 1987. Equivalent Temperature Index at temperatures above the thermoneutral zone for lactating dairy cows. ASAE, Paper 87-4015. St. Joseph MI: USA.

Berry, I. L., M. D. Shanklin and H. D. Johnson. 1964. Dairy shelter design based on milk production decline as affected by temperature and humidity. Transactions of the ASAE. 7(3): 329-331.

Buffington, D. E., A. Collazo-Arocho, G. H. Canton and D. Pitt. 1981. Black globe-humidity index (BGHI) as comfort equation for dairy cows. Transactions of the ASAE. 24(4): 711-714.

Costa Neto, P. L. O. 1977. Estatística (Statistics), São Paulo, Brazil, Edgard Blücher Editor Ltd., 64p.

Johnson, R.A., Wichern, D.W. Applied Multivariate Statistics, Prentice Hall, 1998

Naas, I. A., C. Lagana , D. J. Moura, L. A. Laloni, M. A. Aguiar, R. L. S. P. Reis. 1993. Modelo Matematico para Producao de Leite (Mathematical model for milk production). Proceedings of the XIII Brazilian Congress of Animal Production, Brasilia - DF. p 85-91.

Naas, I. A., D..J. Moura, K.V. Sydenstricker. 1991. Evaluation of Dairy Housing Under Tropical Climate. Asae. Paper 91-4024. St. Joseph, Mi.

Naas, I. A., I. J. O. Silva, D. J. Moura, V. R. C. Carvalho, R. A. Bucklin And F. S. Zazueta. Model validation for semi intensive milking cows production. Proceedings XII CIGR World Congress and Conference on Agricultural Engineering, Milan - Italy. August, 1994. Abstract p. 285-286. Paper No. 94-C-056.

Thom, E.C. 1958. Measuring the need for air conditioning. Air Conditioning, Heating and Ventilation. 53(8): 68-70.