Dietary grouping strategies to improve profitability on dairy farms

Grouping cows is a common practice farmers use to manage their herd more efficiently. Farmers use this strategy to separate far-off, close-up, sick, fresh, and pregnant cows. Grouping addresses cows’ specific needs. However, grouping lactating cows for nutritional purposes and providing groups with more precise diets has not been adopted as widely as it could be, despite the fact that many studies demonstrate its economic benefits (McGilliard et al., 1983; St-Pierre and Thraen, 1999; Cabrera et al., 2012). Reasons of farmers not favoring nutritional grouping can be attributed to farm physical limitations (such as machinery and facilities), labor cost, difficulty in managing multiple diets, and the presumption that milk production will be depressed due to pen or group changes.

Total mixed rations (TMR) have become an industry standard and a large number of dairy farms are using just one TMR for all lactating cows. Applying just one TMR to all lactating cows results in more over conditioned cows and higher nutrient excretion. These alone could be enough reasons to consider additional nutritional grouping as a valid dairy herd management option.

Moreover, adopting additional nutritional feeding strategies could substantially decrease feed costs that represent the single largest expense for dairy production and determine largely herd profitability. Additional nutritional grouping strategies will therefore increase the herd income over feed costs and contribute largely to improved profitability and economic efficiency.

This simulation study aims to evaluate and demonstrate the economic and environmental benefits of nutritional grouping strategies on lactating dairy cattle. Furthermore, this paper also includes a description of a simplified online decision support tool that could be used to evaluate farm-specific grouping strategies for feeding lactating dairy cattle.

 

A Monte Carlo stochastic and dynamic model was developed to simulate each individual cow of a herd to study different nutritional grouping strategies. The model mimicked cows’ life events within reproductive cycles. An individual cow reproductive cycle started by calving and ended by involuntary culling, death, or entering a next lactation. The model used the next event scheduling approach (De Vries, 2001). These events included: involuntary culling (and reasons for culling), death, pregnancy, abortion, dry-off, and parturition. For each event, a two-step approach was followed: 1) determining the binary outcome of the event (it happens or not) and if it happens 2) the day of the occurrence.

After scheduling all the events, the status of every cow in the herd was updated on a daily basis according to scheduled events. For example, if a cow was scheduled to be pregnant at 90 DIM, then the cow information was updated as pregnant when 90 DIM. When the event involved culling or mortality, replacement was assumed to occur the next day to maintain the herd size constant (Cabrera, 2012). Replacement occurred with a springer heifer at the time of first calving.

The model was then used to evaluate the economic and environmental impacts of nutritional grouping strategies. Under this framework, differences in outcomes between dietary grouping strategies were deemed to be more important and relevant than the absolute predicted values.

The model starts with actual data from a commercial dairy herd. It first reads an input file containing cowlevel information of a herd. Then, it performs a projection of each cow and of the whole herd over a year. Parameters included in the input file are: Cow id, parity, DIM, days in pregnancy, fat percentage, protein percentage, 305 mature equivalent (ME305 in kg/cow), body weight (if available), and predicted producing ability (PPA) -producing ability of cow in future lactation- (if available).

The following sections describe the methods used for calculating the attributes of an individual cow on a given lactation and DIM. Some of these attributes were held constant for the lifetime of cow (e.g., PPA), and some varied based on lactation and DIM (e.g., lactation curves).

Cows in the herd were categorized based on their PPA or, in its absence, on their milk mature equivalents at 305 days (ME305). Therefore, every cow was classified based on their future capability of production relatively to categorization by scaling up or down their production. Then, an incomplete gamma function (Wood, 1967) was used to calculate daily milk production of a cow.

Data from Pinedo et al. (2010) were used to determine risks of culling and death and also the reasons for culling. Non-pregnant cows with DIM > 300 were marked as do-not-breed and they were culled (reproductive failure) whenever their milk production reached < 24 kg/d (Kalantari and Cabrera, 2012).

Voluntary waiting period (VWP) was set to 50 d. Cows with DIM ≥ VWP were observed for estrus and had a risk for breeding and conception. Cows detected in estrus and conceiving were marked as pregnant. An 8% risk of pregnancy loss was considered from 30 days to term with an empirical distribution determining the day of occurrence. Gestation length of a cow was modeled using a distribution between 266 and 290 days (Oltenacu et al., 1980).

Cow-specific initial body weight (BW) came from dairy farm records or, if not available, was sampled from a triangular distribution with a minimum of 550 kg, a maximum of 1,000 kg, and a most likely value of 700 kg. An underline Korver function (Korver et al., 1985) was used to simulate BW changes through lactation.

Daily dry matter intake (DMI) was calculated using NRC (2001) equation, a function of maintenance and milk production, adjusted for decreased DMI during the early lactation period.

Total net energy (NE) requirement of a cow was calculated by aggregating the requirements for maintenance (NEm) and milk production (NE_L) based on NRC (2001) equations.

Total crude protein (CP) requirements were also calculated by summing up the CP for maintenance (CP_m) and for milk production (CP_L) based on equations presented in McGilliard et al. (1983).

Milk price was set to $0.35/kg of milk. The nutrients costs were set to $0.116/Mcal of NE and to $0.747/kg of CP based on Cabrera et al. (2012).

Cows were assigned to 2 types of groups: 1) obligated groups and 2) optional groups. Obligated groups included dry cows and fresh cows (1 to 21 DIM). Optional groups were used to test the effect of nutritional grouping on the overall income over feed costs (IOFC), nutrient excretion, and efficiency of the nutrient used by cows.

Cows were ranked and grouped based on actual CP and NE_L requirements (cluster grouping, McGilliard et al., 1983). Every month, cows were regrouped and reassigned to the same or different group based on their rank. Allocation of cows to optional groups was designed to maximize the IOFC (Cabrera et al., 2012). An effect of milk depression of 1.82 kg/day for 5 days for cows moving to a new group was included.

Diet formulations were based on group of cows’ requirements of NE_L and CP and were set at the 83rd percentile (Stallings and McGilliard, 1984), but based on actual group nutrient requirements and not on group milk production.

Five dairy herds were studied (Table 1) according to nutritional grouping strategies. Analyses included calculations of economic efficiency and energy and N use efficiencies because of nutritional grouping. Analyses included the effect of milk depression on the economic and nutrient use efficiency.

 

Table 1. Dairy herds analyzed using the stochastic dynamic model

 

Based on principles above described, a simplified online decision support tool has been developed, grouping strategies for feeding lactating dairy cattle, and it is openly available at the University of Wisconsin-Madison Dairy Management Website (http://DairyMGT.info: Tools). Briefly, users can perform herd-specific analyses after entering cow-level records. Users need to enter cow-level records, define current herd nutritional characteristics, and describe potential changes related to nutritional grouping. The tool calculates the economic difference between current and proposed situations (Cabrera et al., 2012).

There are 2 main structural differences between the online decision support tool and the research methodology described above. The first one is related to the stochasticity. The decision support tool does not contain stochastic elements in the calculations. The second major difference is related to dynamics. Whereas the methodology described above simulates day-by-day every cow on the herd, the decision support tool performs the analysis at one specific point in time (based on records entered by user). In consequence, the decision support tool does not perform projections and therefore should be used frequently (i.e., every month). The decision support tool demonstrates the economic value of nutritional grouping and it is better suited for initial approximated extension messages. Advantage of the decision support is its user-friendliness and the opportunity of interactivity for “what if” analyses. Also, the decision support tool counts with an additional module that can evaluate additional costs or savings beyond milk income and feed costs when performing nutritional grouping management strategies.

Cow-level records of 30 Wisconsin dairy herds were analyzed using the tool. Without knowing their actual nutritional grouping strategies, same procedure was applied at each one of the herds: 1) comparisons were always between 1 group and 3 same-size groups; 2) prices were set at milk ($0.35/kg), CP ($0.3158/kg), and NE_L ($0.1174/Mcal) for all farms; 3) requirements of CP and NE_L were at the 83rd percentile for both 1 or 3 groups; and 4) average BW was set at 500 kg for primiparous and at 590 for multiparous.

 

Economic efficiency measured as IOFC increased consistently with additional nutritional groups (Fig. 1). On average and compared to 1 nutritional group, the IOFC gain ($/cow per year) was 21, 46, 57, 65, and 70 for 2, 3, 4, 5, and 6 nutritional groups, respectively. Curves in Fig. 1 follow a decreasing increasing trend (law of diminishing returns) after 3 nutritional groups, indicating that there is still a gain for more than 3 nutritional groups, but the gains decrease. It is also important to notice that there is a relationship between number of groups and herd size, indicating a higher opportunity of grouping in larger herds. In all cases, more than 4 or 5 nutritional groups might be impractical, but are still provided for analyses purposes. In small herds (e.g., 331), more than 3 groups might be impractical.

 

Figure 1. Economic efficiency of nutritional grouping strategies in 5 Wisconsin dairy herds (Table 1). Labels indicate the number of lactating cows on herd

 

Utilization of energy measured as the percentage of energy in milk (Mcal milk) over the energy consumed in feed (Mcal consumed) increased consistently with additional nutritional groups until 4 groups (Fig. 2). It continued increasing for the largest herd (1460) until 6 groups and for the 727 herd until 5 groups. This indicates clearly that grouping strategies are herd specific and depend on herd size. On average and compared to 1 nutritional group, the efficiency of energy use increased (%) 0.18, 0.59, 0.81, 0.82, and 0.87 for 2, 3, 4, 5, and 6 nutritional groups, respectively (Fig. 2). It is interesting to notice that in the largest herd (1460) there was no apparent gain from 1 to 2 groups and the gain from 2 to 3 groups was smaller than other herds. Nonetheless, this herd showed greater gains than other herds with more nutritional groups.

Utilization of N measured as percentage of N in milk (kg) over the N consumed in feed (kg) also increased consistently with additional groups (Fig. 3).

 

Figure 2. Energy use efficiency of nutritional grouping strategies in 5 Wisconsin dairy herds (Table 1). Labels indicate the number of lactating cows

 

Differently than energy, with exception of herd 570, N efficiency continued to increase up to 6 groups (at a diminishing return). On average and compared to 1 nutritional group, N efficiency (%) was 0.36, 0.31, 0.17, 0.07, and 0.06, for 2, 3, 4, 5, and 6 nutritional groups, respectively.

 

Figure 3. Nitrogen use efficiency of nutritional grouping strategies in 5 Wisconsin dairy herds (Table 1). Labels indicate the number of cows

 

Milk depression because of cows being relocated to different groups, as defined in this study (1.82 kg/day milk loss for 5 days = 9.1 kg milk), had an overall economic impact that decreased IOFC between $16 (2 nutritional groups) and $23 (6 nutritional groups) on herd having 787 lactating cows (Fig. 4, difference between bars). This strategy also had an overall N efficiency use impact that decreased the ratio milk N produced/feed N consumed between 0.05% (2 nutritional groups) and 0.13% (6 nutritional groups) (Fig. 4, difference between curves).

 

Figure 4. Economic and N use efficiency with and without considering milk depression on the herd 787 (787 lactating cows).

 

Considering that the IOFC gain for 2 nutritional groups without milk depression would have been $33, the impact of milk depression represents a 48% decrease in the gain just because of milk depression, a large impact. The impact of milk depression on IOFC decreased as more nutritional groups were in place. This was about 30% for 3 nutritional groups. Also, considering that the IOFC gain for 6 nutritional groups would have been $90, milk depression represents a 25% decrease in the gain, still an important impact.

Evaluations of 30 Wisconsin dairy farms with online decision support tool (http://DairyMGT.info: Tools: Grouping strategies for feeding lactating dairy cattle) consistently demonstrated in all 30 herds that IOFC was greater for the nutritional grouping strategy that included 3 nutritional groups compared with the strategy of only 1 nutritional group (Table 2).

 

Table 2. Economic nutritional grouping evaluation of 30 Wisconsin dairy farms using the online tool Grouping strategies for feeding lactating dairy cattle

 

The analysis indicated that farms could realize between $161 and $580/cow per year (mean = $396) of additional IOFC by switching from no grouping to 3 same-size feeding groups using the cluster criterion for grouping. These values represented an increase of between 7 and 52% of farm calculated IOFC.

Performing grouping and feeding different rations to the groups could have additional costs and possible economic losses. After assuming reasonable costs of management, labor, and machinery and reasonable expected milk depression (1.82 kg/day for 5 days) on those cows affected by the grouping changes, the net return of grouping was still much greater than the no groping option. The additional IOFC estimated in Table 1 decreased between 9 and 25% for these scenarios. Therefore, 3 same-size feeding groups was still much more profitable in all 30 herds than the sole 1 feeding group option.

 

We conclude that additional nutritional grouping on lactating dairy cattle has a positive impact on the economic and environmental efficiency of a dairy farm. Benefits will vary largely depending on farm and market conditions, but our conservative projection analyses indicate that a farm could expect about $45/cow more IOFC per year, 0.59% increased energy efficiency use, and 0.31% increased N efficiency use when a farmer would decide to manage 3 nutritional groups considering that currently is managing only 1 nutritional group for all lactating cows. A user-friendly online decision support tool is openly and freely available at the University of Wisconsin Dairy Management Website (htttp://DairyMGT.info) and could be an effective tool for initial demonstration and motivation to promote nutritional grouping on lactating dairy cattle.

 

 

This is an updated paper presented at the Western Canadian Dairy Seminar, Red Deer, Alberta, 11-14 March 2014.

Funding for this project was provided by Agriculture and Food Research Initiative Competitive Grant no. 2011-68004-30340 from the USDA National Institute of Food and Agriculture.

 

 

 

 

1. Cabrera, V. E. 2012. A simple formulation and solution to the replacement problem: A practical tool to assess the economic cow value, the value of a new pregnancy, and the cost of a pregnancy loss. Journal of Dairy Science 95:4683-4698.

2. Cabrera, V. E., F. Contreras, R. D. Shaver, L. E. Armentano. 2012. Grouping strategies for feeding lactating dairy cattle. Pp. 40-44 in Proceedings Four-State Dairy Nutrition and Management Conference. Dubuque, IA, 13-14 June 2012.

3. De Vries, A. 2001. Statistical process control charts applied to dairy herd reproduction. PhD Thesis. University of Minnesota. Minnesota.

4. Kalantari, A. S., and V. E. Cabrera. 2012. The effect of reproductive performance on the dairy cattle herd value assessed by integrating a daily dynamic programming with a daily Markov chain model. Journal of Dairy Science 95:6160–6170.

5. Korver, S., J. A. M. van Arendonk, and W. J. Koops. 1985. A function for live-weight change between two calvings in dairy cattle. Animal Production 40:233-241.

6. McGilliard, M. L., J. M. Swisher, and R. E. James. 1983. Grouping lactating cows by nutritional requirements for feeding. Journal of Dairy Science 663:1084-1093.

7. NRC. 2001. Nutrient requirement of dairy cattle. 7th rev. ed. National Research Council, Washington, DC.: National Academy Press.

8. Oltenacu, P. A., R. A. Milligan, T. R. Rounsaville, and R. H. Foote, 1980. Modelling reproduction in a herd of dairy cattle. Agricultural Systems 5: 193-205.

9. Pinedo, P. J., A. De Vries, and D. W. Web. 2010. Dynammics of culling risk with disposal codes reported by dairy herd improvement dairy herds. Journal of Dairy Science 93:2250-2261.

10. St-Pierre, N. R., C. S. Thraen. 1999. Animal grouping strategies, sources of variation, and economic factors affecting nutrient balance on dairy farms. Journal of Animal Science 77:72-83.

11. Stallings, C. C., and M. L. McGilliard. 1984. Leading factors for total mized ration formulation. Journal of Dairy Science 67:902-907.

12. Wood, P. D. P. 1967. Algebraic model of the lactation curve in cattle. Nature 216: 164-165.