Volume 3 No. 1 JANUARY 2012 • pages 7-13 Malaysian Journal of Veterinary Research

A BEEF FATTENING DECISION SUPPORT SYSTEM

SHANMUGAVELU S.1, WAN ZAHARI M.2, WONG H.K.1 AND MARDHATI M.1 1 Malaysian Agricultural Research and Development Institute, MARDI, Serdang, Selangor. 2 Universiti , Pengkalan Chepa, , Kelantan. [email protected]

ABSTRACT. A beef feedlot production INTRODUCTION decision support system (DSS) was developed based on Microsoft® Excel. Beef is an important commodity in The DSS comprises of three modules i) an Malaysia with a per-capita consumption ingredient database ii) a least-cost ration of 5.6 kg. However, only 28% of this formulation module and iii) beef growth requirement is produced locally (DVS, simulation module. The program uses 2010). The main factor that contributes empirical equations developed for tropical to the low self sufficiency level is the beef to simulate nutrient requirements high cost of local beef production. For and daily body weight gains based on the example, the average cost of local beef for formulated feed ration. The formulated 2010 was RM15.85 compared to RM9.20 least cost ration can be pasted automatically for imported beef from India (DVS, into the growth model to evaluate 2010). The lack of cheap feed and the performance and economic viability. The inefficient use of available feed resources growth model calculates nutrient available contribute significantly to the higher cost and computes body weight gain on a daily of production as feed generally comprises basis, summates weight gain and stops 60-70% to the total production cost. Many at the targeted body weight. The data local beef producers do not have access to output include i) days to reach target body information on nutrient values of available weight, ii) cumulative feed consumed, iii) feed resources nor the ability to efficiently anticipated average daily gain, iv) total utilise the resources. This paper describes cost of feed (concentrates and grass), and a beef fattening decision support system v) gross profit per cattle. If a portion of that can help improve the efficiency of a the feed is fed as grass, then the model beef production enterprise. also computes the pasture land required in hectares, based on the forage species MATERIALS & METHODS chosen. It is anticipated that the developed model can assist cattle entrepreneurs and A Microsoft® Excel based software was farmers in the development of the beef developed based on cited publications cattle industry in Malaysia. (NRC, 2000; Leonard, 1982) and beef growth data collected from research conducted in MARDI. The model comprises

7 Malaysian Journal of Veterinary Research Volume 3 No. 1 JANUARY 2012

Least Feed Cost Database Module

Simulation module

Cattle Select Initial Target Cattle No. of Sale Feed Body wt Body wt Purchase Cattle (RM/kg) (RM/kg)

Model Inputs

If Body Weight < Target Weight Analyze Body Weight If Body Weight = Target Weight

Output

1. Total feed cost (grass + concentrate) 2. Monthly gross profit & profit/cattle 3. Grass land required 4. Average daily gain (kg) 5. Dry matter intake

Figure 1. Model algorithm

of three modules i) an ingredient database based on research data to simulate ii) a least-cost ration formulation module nutrient requirements and daily body and iii) beef growth simulation module. weight gains. The nutrient requirements The ingredient data base comprises of for the beef production module are based nutrient content information of local feed on Department of Standards Malaysia resources. The least cost module utilizes (Standards Malaysia) (unpublished). the linear optimization module inbuilt in Microsoft® Excel 2007 Visual Basic for Microsoft® Excel 2007 for Windows. The Application was employed for the beef beef growth simulation uses empirical daily growth simulation module. The equations developed for tropical beef program algorithm is shown in Figure 1.

8 Volume 3 No. 1 JANUARY 2012 Malaysian Journal of Veterinary Research

Other features incorporated into the model RESULTS AND DISCUSSION include the options to evaluate two different feed formulations and their growth rate The captions of the beef fattening decision predictions. support system are shown in Figures 2, 3, 4, 5, 6, 7 and 8 with simple user friendly

Figure 2. Main Screen

Figure 3. Growth simulation module

9 Malaysian Journal of Veterinary Research Volume 3 No. 1 JANUARY 2012

modules. The model enables the user to two types of formulations (Figure 5). The alter feed ingredients and test different feed user has the option to alter, add and make combinations based on costs and nutrient changes to the feed ingredient database requirements and allows the comparison of including the use of grass as a proportion

Figure 4. Ingredient database

Figure 5. Least cost formulation module

10 Volume 3 No. 1 JANUARY 2012 Malaysian Journal of Veterinary Research

of the diet. If grass is chosen as an option, beef fattening decision support system was the model computes the land required to verified against actual beef cattle growth cultivate the grass species chosen. The from studies conducted in MARDI.

Figure 6. Least cost comparison module

Figure 7. Model evaluation module

11 Malaysian Journal of Veterinary Research Volume 3 No. 1 JANUARY 2012

Figure 8. Model analysis and economic evaluation module

220

210 Actual 200 Predicted g ) k (

190 g h t i

e 180 w

170 o d y

B 160

150

140 0 14 28 42 56 70 84 98 112 Age (days)

Figure 9. Actual vs Predicted body weight

12 Volume 3 No. 1 JANUARY 2012 Malaysian Journal of Veterinary Research

A sample data of predicted against DSS software can be utilised to optimise actual growth rate based on the feed returns. The model can be used by feedlot formulated is shown in Figure 9 with an operators, beef nutritionists, and also in average prediction error of 0.8 kg. The universities as a teaching tool. However, model developed can be used by the it is emphasised that as with most DSS beef feedlot industry to make intelligent systems, there can be variations in any decisions and avoid losses in feedlot biological system. Nevertheless this operations. It can also be used by extension software can be used as an intelligent tool agents and as a teaching tool especially in to assess feedlot operations and improve universities. economic returns.

CONCLUSION REFERENCES

1. DVS (2010). Online Livestock Statistics. Department of A beef feedlot production decision Veterinary Services, Malaysia. http://www.dvs.gov.my/ support system (DSS) developed based on web/guest/perangkaan. Accessed 23 May 2011. 2. NRC (2000) Nutrient Requirements of Beef Cattle. Microsoft® Excel was observed to predict Seventh Revised Edition 1996. National Academy Press, beef growth under Malaysian conditions Washington, DC, USA. 3. Leonard C.K. (1982). Nutrient Requirements of ruminants within reasonable limits. Beef feedlot is in developing countries. International Feedstuffs a challenging enterprise especially with Institute, Utah State University, Logan, Utah, USA. the high cost of feed ingredients and this

13