[Skip to Content]

Illinois Livestock Trail
NOTICE - This information has been archived and may contain outdated content.
Feed Intake Behavior and Implications for Energy Utilization
by Mike Ellis and Young Hyun

The study of feed intake has been greatly facilitated by the recent development of computerized systems that monitor the feed intake of individual animals within a group. Such equipment was originally designed for use by breeding stock supply organizations in the performance testing of pigs in group situations. The majority of selection programs for pigs place considerable emphasis on feed efficiency because of its impact on production costs. Historically, genetic improvement in feed efficiency was achieved as a correlated genetic change resulting from selection for growth rate and carcass lean content for animals tested in groups, where individual feed intakes could not be measured. Alternatively, pigs were penned and fed individually to obtain feed intake data, a procedure that is expensive in terms of both labor and facility costs. In addition, animals that are genetically superior when penned as individuals may not perform the best in group-housed situations i.e. under commercial conditions. Studies that have compared the growth and carcass characteristics of boars that have been individually fed during the performance test with the performance of their progeny under commercial group-housed conditions. These have generally shown moderate to low correlations between performance in the two environments. This phenomenon is termed a genotype x environment interaction and suggests that different genes may be involved in controlling growth performance in group-versus individually housed animals. Thus, breeding organizations stimulated the development of feed intake recording equipment to allow them to test and select animals in a social environment similar to that experienced in commercial production.

Computerized feed intake recording systems are generally used with growing-finishing pigs and they have the capacity to take pigs between live weights of approximately 20 to 150 kg. They comprise a feed station, which consists of a single-space feed trough attached to a load cell, an antenna that picks up signals from the individual pig transponders, control equipment and a computer which collect and store all of the data on feeding activity at the feed station. The feed station generally has some form of a protective race which ensures that only one pig can access the feed trough at any time. The major difference between the various systems that are commercially available is in the design of the protective race, which vary from low to high levels of protection for the feeding pig.

When a pig enters the feed station, the signal from its ear tag transponder is picked up and the identity of the animal is recorded, together with the time of the visit and the weight of the feed trough. The same information is recorded as the animal leaves the feed station and the amount of feed consumed and the time spent in the feeder is obtained by difference. Data on feed consumption at each visit is accumulated over time to give daily and total test period feed intakes for each animal in the group.

A number of commercial designs of computerized feeders have been developed including the following:

ACEMA - 48 (Centre d’Etudes du Machinisme Agricole, du Genie Rural et des Eaux et Forets,

F-35000, Rennes, France)

Feed Intake Recording Equipment (FIRE, Hunday Electronic Ltd, Newcastle upon Tyne,

England; Osborn Industries, Osborn, Kansas).

Individual Voluntary Feed Intake Recording in Group Housing (IVOG, Technical and Physical Services in Agriculture, Wageningen, Netherlands)

Tedea Modal 1250 (Tedea, Israel; Eweka Electronic Identification System, Pittsburg).

This later system is designed principally for research use with the other three systems being designed largely for use for performance testing breeding stock to obtain data on the total amount of feed consumed by individual animals. However, they also provide a considerable amount of detail on the feeding activity of animals in groups including data on the following:

  • the time and duration of visits to the feeder
  • the amount of feed consumed at each visit
  • the feed consumption rate (which has generally been defined as feed consumed ÷ time in the feeder).

This equipment, therefore, has the potential to be a valuable research tool which can be used to provide an insight into feeding behavior. A FIRE system was installed at the University of Illinois in 1993 and the results of our studies are presented, along with research from other centers, in this paper.

An interesting feature of computerized feed intake recording equipment is that it generates a huge amount of data relative to feed intake. The results of a recent study carried out at this center involving 120 pigs over a 10 week period illustrate this point (Hyun et al., 1996). Animals visited the feeder on average 12 times per day and for each visit the time of entry and exit and the weight of the trough before and after each visit in the feeder were recorded directly. From this information, the time spent in the feeder and the amount of feed consumed at each visit are calculated, thus producing an average of over 5000 observations for each individual pig and more than 700,000 total observations for the study relating to feeder activity.

A recent development has been to addition of a weighing platform to the feed station which allows the weight of animals to be automatically recorded at the same time as their feed intake. From the standpoint of genetic improvement programs, it is now theoretically possible to select pigs that have specific growth and feed intake curves. For example, it may be possible to select animals that have higher feed intakes during the growing phase and lower feed intakes during finishing.

Despite the widespread use of this type of equipment by the seed stock industry, there is surprisingly little information on the most appropriate conditions for its use, particularly in terms of group sizes and space allocation. Manufacturers’ recommendation have been to allow relatively large amounts of space per animal, up to twice the normal recommended levels for commercial pigs. However, this recommendation appears to be related more to reducing ear tag loss than to any consideration of the impact of space allocation on performance levels or feeding behavior. Early versions of the FIRE system used relatively large ear tag transponders which were subject to a high level of losses. However, more recent versions use a much smaller transponder which has much lower loss rate.

We have recently completed a study on the FIRE system investigating the influence of group size of 2, 4, 8 and 12 pigs per pen for animals between 25 kg and 50 kg live weight, at a stock density 0.9 m²/pig (Hyun and Ellis, 1996). The results of this study are summarized in Table l and Figure l. Daily feed intake and live weight gain decreased with group size, largely as a result of a reduction in the total amount of time spent in the feeder each day; pigs in groups of 12 spent approximately 24 and 29 minutes less time in the feeder and consumed 0.20 and 0.16 kg less feed per day than groups of 2 and 4 pigs respectively (Table l). The diurnal distribution of feeder visits (Figure la) was characteristically higher during the daytime (0600 to 1800 h) than during the nighttime. However, feeder activity for the groups of 12 pigs showed much less diurnal variation than for the smaller groups and the percentage of time that the feeder was occupied increased with group size and approached 100% for most of the 24 h period in the largest groups (Figure 1c). Further research is in progress to investigate the impact of group size on growth and feed intake traits in finishing pigs.

The major difference between the FIRE system and conventional commercial feeders is the presence of the protective race in front of the feed hopper. Nielsen et al. (1995b) investigated the effect of three different entrance designs on growth performance and feeding behavior in groups of 10 pigs between 34 and 57 kg live weight. The designs compared were a short race (head-guard), a standard full-length race (similar to the race used with the FIRE system in studies at the University of Illinois), and a full-length race with a pneumatic gate. The authors characterized these races as affording low, medium, and high protection against disturbance of the feeding pig respectively. The results of this study, which are summarized in Table 2, suggest small difference in performance and feeding behavior between pigs on the system with the head-guard and the standard full-length race. However, animals using feeders equipped with the protective-race with the gate had fewer but longer feeding visits compared to the other two treatments (Table 2). This study suggests that using a protective race such as has been used in the University of Illinois studies will have minimal impact on performance and feeding behavior compared to designs of race that provide less protection.

Feed Intake Behavior

Computerized feed intake recording systems have been available commercially for less than 10 years and, therefore, research using such equipment has only been published over the last 5 years or so. To date most of the published research has been carried out in Europe, particularly in Holland, France and the United Kingdom. A summary of research results published to date using computerized feed intake recording systems is presented in Table 3. This table illustrates the tremendous variation in the results of studies in feed intake behavior traits. For example, the number of daily feeder visits range from as low as 7 to as high as 73 with this range encompassing group sizes of 20 and individually penned pigs, respectively. Similarly, feed intake and feeder occupation time per visit range from 35 g and l.4 min to 222 g and 6.9 min, respectively. Feed consumption rate, which has generally been defined as amount of feed consumed divided by the time spent in the feeder, ranged from as low as 15 g/min up to as high as 40 g/min (Table 3).

Growing-finishing pigs show a characteristic diurnal distribution in feeder activity an example of which is illustrated in Figure 2, which is taken from the study of Hyun et al. (1996). Visits to the feeder and feed consumption were greatest during the daytime, showing a characteristic peak between approximately 0600 h and 2000 h, and lowest during the nighttime (Figures 2a and 2b). The behavior of pigs in the feeder also showed a diurnal distribution with those pigs that visit the feeder during the peak in activity having shorter visits (Figure 2e), eating less per visit (Figure 2d), but having higher consumption rates (Figure 2f) than animals visiting the feeder during the nighttime. In this study (Hyun et al, l996), the three sexes (boars, barrows and gilts) were compared and the results presented in Figure 2 suggest that the differences between barrows and gilts in feeding behavior were small. However, boars appeared to have more of their feeder visits and consume a bigger proportion of their total feed between 0700 h and 1600 h when activity at the feeder was greatest. This suggests that the boars were more successful at gaining access to the feeder when competition is highest.

Evidence of genetic variation in feeding behavior is shown in a small study carried out at this center by Hyun et al. (l996a) which compared purebred Yorkshire and Meishan barrows reared in mixed-breed groups of 10 pigs. The growth performance and feeding behavior of the two breeds was dramatically different (Table 4). The daily feed intake of the Meishans was approximately 1 kg lower than that of the Yorkshires and the Meishans made less than half the number of feeder visits but had a greater feed intake per visit compared to the Yorkshires (Table 4). In addition, the diurnal pattern of feeding behavior differed between the two breeds with Yorkshires consuming proportionately more of their feed at times of peak feeder activity (Figure 3b). It appeared that the Meishans were reluctant to compete for feeder access and, therefore, visited the feeder after the Yorkshires had consumed the bulk of their daily feed. Thus, a response of subordinate pigs in a group to increased competition for feeder access appears to be to eat at times of the day outside of the normal peak in feeding activity.

Individual Variation in Feeding Behavior

One of the striking features of daily feed intake behavior data is the tremendous variation observed both between individual animals and between consecutive days for the same animal. For example, the range in mean number of feeder visits per day for individual pigs of the same genotype, sex and weight was from 8 to 23 in the study of Hyun et al. (1996) and even greater (from 3 to 69 visits/day) in the study of Young and Lawrence (1994). Day to day variation in feed intake level in individual pigs is illustrated in Figure 4 where pigs from the study of Hyun et al. (1996) with high and low daily variation have been selected. It has been proposed that under commercial conditions feed intake could be measured on a daily basis and used to predict future feed intake. This information would be of value to producers in diet formulation. The data presented in Figure 4 raises questions about how accurate such as approach might be.

Feed Intake Behavior and Energy Utilization

Any impact of variation in feeding behavior on the energy balance of the animal is likely to operate either through an increase in energy supply to the animal or through a reduction in energy utilization by the animal. There is evidence from Dutch research of a relationship between feeding behavior and nutrient digestibility. Thus De Haer and De Vries (1993a) found that 86% of variation in protein and dry matter digestibility was explained by variation in feed intake pattern in individual housed pigs. Correlations suggested that pigs that had a higher number of short visits and ate a lower proportion of large meals had higher digestibility coefficients. However, with animals in groups (8 pigs per group) the correlations between feed intake pattern and dry matter and crude protein digestibility were much lower and variation in feeding pattern accounted for only 20% of the variation in digestibility.

The impact of variation in feeding behavior on energy utilization has not been researched. It is possible, however, that pigs that have higher levels of feeder related activity will expend more energy in this respect. However, it is uncertain if their total energy expenditure will be any different from those with lower levels of feeder activity. Our studies have generally found low correlations between feeding patterns and growth rate, feed efficiency and carcass composition (Table 5). This suggests that, overall, any impact of feeding behavior on energy balance, is likely to be minimal under commercial conditions.


De Haer, L.C.M. and A.G. De Vries. 1993a. Feed intake patterns of and feed digestibility in growing pigs housed individually or in groups. Livestock Production Science 33:277.

De Haer, L.C.M. and A.G. De Vries. 1993b. Effects of genotype and sex on feed intake pattern of group-housed growing pigs. Livestock Production Science 36:223.

De Haer, L.C.M., J.W.M. Merks, H.G. Kooper, G.A.J. Buiting and J.A. van Hattum. 1992. A note on the IVOG-station: A feeding station to record the individual food intake of group-housed growing pigs. Animal Production 54:160.

De Haer, L.C.M. and J.W.M. Merks. 1992. Patterns of daily food intake in growing pigs. Animal Production 5: 94.

Feddes, J.J., B.A. Young and J.A. De Shazer. 1989. Influence of temperature and light on feeding behavior of pigs. Applied Animal Behavioral Science 23:215.

Hyun, Y. and M. Ellis.1996a. Effect of genotype on feed intake pattern and growth performance in group-housed growing pigs. Journal of Animal Science 74(1):51.

Hyun, Y. and M. Ellis.1996b. The effect of group size on growth performance and feed intake level and pattern in growing pigs. Journal of Animal Science 74(1):179.

Hyun, Y., M. Ellis, F.K. McKeith and E.R. Wilson. 1996. Feed intake patterns of group-housed growing-finishing pigs monitored using a computerized feed intake recording system.  Journal of Animal Science (In press).

Labroue, F., R. Gueblez, P. Sellier, and M.C. Meunier-Salaun. 1994. Feeding behavior of group-housed Large White and Landrace pigs in French central testing stations. Livestock Production Science 40:303.

Nielsen, B.L., A.B. Lawrence and C.T. Whittemore. 1995a. Effects of group size on feeding behavior, social behavior, and performance of growing pigs using a single-space feeder. Livestock Production Science 44:73.

Nielsen, B.L., A.B. Lawrence and C.T. Whittemore. 1995b. Effects of single-space feeder design on feeding behavior and performance of growing pigs. Animal Sciences. 6l:575.

Xin, H. and J.A De Shazer. 1992. Feeding patterns of growing pigs at warm constant and cyclic temperatures. Transactions of the ASAE. 35:319.

Young, R.J. and A.B Lawrence. 1994. Feeding behavior of pigs in groups monitored by a computerized feeding system. Animal Production 58:145.

Tables and figures available in PDF version.


« Back to Swine