Introduction
Controlling production costs and animal welfare are important factors for breeders. Feed accounts for 50 to 70% of the production cost in poultry production (van Horne, 2018). The feed conversion ratio (FCR), which is the ratio of feed consumption to meat or egg production, is an important criterion for the profitability of these farms. It is also an indicator of the environmental impact of production. For example, the reduction of 10% in FCR results in a reduction in the production cost of live chicken of about 6%, but also a reduction in environmental impact of 12% for nitrogen excretion and 17 % for phosphorus (De Verdal et al., 2011). So far, individual measurements of feed efficiency and feeding behavior have been carried out on animals reared in individual cages, but this method shows several drawbacks, including lack of social interactions with other animals and movements limitations, which leads to experimental bias, as conditions are not representative of actual production conditions. The bias is particularly important in the case of alternative productions such as slow-growing lines, which have more physical activity. In addition, there is a need of individual data for geneticists to make progress in genetic selection, by being more accurate and comprehensive on feed efficiency and having the possibility to integrate data relating to feed intake dynamics. This technology could enable the study of the impacts of nutritional and technological characteristics of feed on birds feeding behavior and performances. Access to daily performances would make the study of critical time point such as feed transitions easier. In addition, access to individual data would make it possible to obtain greater statistical power from tests with a much smaller number of animals. This is why research were initiated in 2005 by INRAE to develop a tool to measure individual broilers feed consumption of broilers, leading to a patent registration (Chabault et al., 2013). Since then, various improvements have been made to the feeder, such as automatic data recording and processing, better ergonomics and constructions materials, and the addition of systems for animal weighing (Guettier et al., 2020; Mika et al., 2021). Therefore, the purpose of this paper will be to briefly describe the electronic dispenser and its use, and to give an example of application.
Description of the electronic feed station
BIRD-e has a circular shape with adjustable feet for leveling and stability of the automaton. A cylindrical PVC outer shell protects the internal elements. The inner unit is vertically movable thanks to an electric column fixed to the base, in order to adapt the feeder height to the animals’ size.
The feeder is equipped with eight independent feed accesses, without corridors, so that the chickens can express their natural feeding behavior. Each access includes one feed tube, one feed trough, one antenna placed on the top of the feed trough to detect the animal’s RFID chip, one scale for recording feed weight, and one scale for recording animal weight placed under the tray on which the animal would climb to eat (Figure 1). One feed station is suitable for 100 to 120 animals without causing competition.
Feed is distributed through eight independent columns (1 column for each access). Each column has its own load cell. The measuring capacity of the load cells for the feed is 10 Kg. Each column has a storage volume of 5.6 dm3 and can fill about 3 Kg of feed. The lower part of the feed columns is the feed access for the animals. To avoid waste, the feed access is covered with a transparent cover through which the animal passes its head to eat.
Animals are weighed using interchangeable trays according to their growth: they are clipped onto the animal scale in the axis of the eight accesses. Each tray has its own scale. The measuring capacity of the weighting cells for the animal is 30 Kg.
The RFID antennas are fixed on the feeder, at the entrance of each access. They are adjustable in height, and in the axial position by means of the sliding support and its knurled knob. This system allows the antenna to be positioned as close as possible to the animal's chip throughout the rearing period, which is particularly useful when animals are young.
Figure 1. Description of BIRD-e, the electronic feeder for poultry
Operation of the electronic feeder and data collected
At one day of age, animals are equipped with a RFID (Radio Frequency Identification) electronic chip attached to the base of the neck for individual identification (Figure 2).
Figure 2. RFID chip attached to the base of the neck for individual identification.
When the animal comes to eat, it climbs onto the tray and the chip is identified by the electronic sensor. Sensor is connected to an electronic card that transmits all the identification data to a data acquisition system (PMX, HBM), and then stored in an industrial computer. All computer and electronic equipment are grouped together in the air-conditioned electrical cabinet located above the feeder.
Raw data obtained from the station were 1) feed weight by access every second, 2) animal identity, time, and access number every time an antenna detected a chip, and 3) average animal weight at each visit. A visit was defined by consecutive readings of the same chip at the same access with less than 10 seconds between consecutive chip detections. All scales and antennas were connected to a central data acquisition system.
The identification data is stored in a first file with the following elements: "Date - Time - Feeder number - Access number - Identification number". The animal weight is recorded every time an animal chip is detected. A second data file is generated and contains, for each visit of the animal to the feeder, the "Date - Number of the feeder - Access number – Start time of the visit – End time of the visit - Average weight of the animal during the visit ". The feed weight in each of the feed columns is recorded in real time, second by second, whether an animal is present or not, thanks to the scales attached to each feed column. The resulting file contains one data item per second and per access and the following information is entered: "Date - Time - Feeder number - Access number - Feed weight".
The data acquisition system synchronizes the data from the eight accesses, then transmits it to the central computer, which stores it and then exports it to a server. An algorithm for calculating feed consumption for each visit was developed by determining feed access and access weight at the start and the end of the visit. It allows to determine, for a defined feed access, the total amount of feed consumed per animal. By adding up all the daily feed accesses of a given animal, it is then possible to have access to the amount of feed ingested each day.
In addition to animal performance data (weight, feed intake, feed efficiency), the BIRD-e feeder provides access to new data such as animal feeding behavior. Then, it is possible to know, for each animal, the number of feed accesses made in a day, the time interval between two feed accesses, the average duration of a feed access, the feed intake intensity, etc. This opens up new perspectives for future feeding trials.
Example of application: Study of the interaction between genotype and diet for consumption, feeding behavior and feed conversion ratio (Berger et al., 2021)
In this study, we evaluate the adaptation ability of two genotypes, i.e. fast (FG) and slow-growing (SG) chicken, with different levels of growth rates and nutritional requirements, to a diet containing a mixture of alternative feedstuffs. Birds from both genotypes were reared separately, at different periods. In the first batch, 80 SG male chickens were reared from 1 to 82 days, according to “Label Rouge” rearing practices. In the second batch, 80 FG male chickens were reared from 1 to 35 days. Two different diets were designed, adapted to each genotype nutritional needs: a classic corn-soybean diet as a control (Control) and an alternative diet (Alternative) including more wheat and local rich-protein feedstuffs as sunflower meal and rapeseed meal, and less soybean meal than the control diet. Within a genotype, diets were isoproteic and isoenergetic. At one day of age, animals were identified with a RFID chip for individual identification and were randomly distributed in two pens, each pen receiving either the control or the alternative diet. In each pen, the 40 animals has access to four accesses of the feeder. Daily body weight and feed intake of each animal were continuously recorded throughout the experiment thank to the BIRDe feeder, as well as feed intake dynamics. The kinetics of mean body weight (BW), feed intake (FI), feed efficiency and feed intake dynamics, as well as the variability of these traits between the alternative diet and the control diet were compared from hatch to slaughter.
Within each genotype, BW, FI and FCR of animals receiving either the Control diet or the Alternative diet were very similar, showing the good capacity of animals of both genotypes to adapt to the alternative diet (Figures 3 and 4). We can also see on these figures that, above the mean values, this feed station allows to calculate the coefficient of variation (CV) of each trait, it is thus also possible to check the adaptation to the new feed through the observation of impact on the homogeneity of the group, and not only on the mean value. The CV for BW in both genotypes was stable and low at all ages, usually lower than 20% (Figure 3).
Figure 3. Body weight of animals from SG (A) and FG (B) chickens receiving a control (red) or alternative (blue) diet. Dotted lines represent coefficient of variation of data. Vertical lines represent diet changes. Horizontal green bars indicate significant differences between the two diets. Reproduced from Berger et al. (2021).
However, change with age of daily FI and daily FCR coefficients of variation differed between traits and genotypes, although similar trends were found between diets. The general trend was an increase in the CV of the two traits with age in SG chickens (Figures 2A, 2C) and a decrease in FG chickens (Figures 2B, 2D). Differences of variability between diets for daily FI and daily FCR were strong in FG line. Alternative diet led to a decrease in the variability of those performances during the grower and finisher phases in FG chickens. In the case of SG chickens, the CV differed between diets during these phases for daily FI. When significant, performances were less variable with the Alternative diet than with the Control diet diet.
Figure 4. Feed intake and daily feed conversion ratio from SG (A, C) and FG (B, D) chickens receiving a control (red) or alternative (blue) diet. Dotted lines represent coefficient of variation of data. Vertical lines represent diet changes. Horizontal green bars indicate significant differences between the two diets. Reproduced from Berger et al. (2021).
Regarding feed intake dynamics, SG and FG chickens have on average the same number of feed accesses per day except for FG animals fed the control diet, where higher feed access was observed (Figure 5). However, a significant different in the average feed access duration was observed between the two genotypes, and the two diets only for FG chickens. FG animals have significantly longer feed access than SG animals, which is consistent with their higher feed intake. In the SG line, no difference was observed on this criterion between the diets. However, in the FG line, animals with alternative diet have a longer feed access than animals with the control diet. It can be assumed that this difference between the two diets is due to a different bulkiness, which in turn is due to a higher level of fibers in the alternative diet.
Figure 5. Feed access number and duration for animals from two genotypes and receiving two different diets.
The current application study showed that differences between the two diets are moderate in terms of final performances in both genotypes, indicating that chickens are able to adapt to a diet composed of a mixture of alternative feedstuffs. The most striking difference in the adaptability of chickens to the Alternative diet was found in the variability of performances. Animals fed with Alternative diet had more homogeneous performances for FI and daily FCR, especially in FG chickens. To go further, this design has recently be used to determine genetic parameters of daily data of feed intake and feed efficiency in broilers.
Conclusion
With the BIRD-e feed station, it is possible to conduct experiments under production conditions (on the floor and in groups) and to collect individual measurements of feed intake, body weight and feed intake dynamics. The application example presented here illustrates the interest of this device with the study of different genetics and feeds.
The application study highlighted other interesting aspects of BIRD-e. For example, this device allows us to closely examine consumption during diet changes with a short-term reaction (in relation to sensory changes in diet) and a medium-term reaction (in relation to metabolic effects). The BIRD-e automatic consumption machine represents a real innovation for research and offers new perspectives around genetic selection and animal nutrition. This automaton allows experimentation without altering animal’s behavior and well-being. The impact of this technology is an expected progress in genetic selection, by integrating feeding behavior criteria and being more efficient on feed efficiency. This innovative tool makes it possible to study the impacts of feed composition and technology on the birds feeding behavior and performance. Today, six electronic feed stations have been developed and used by INRAE and ITAVI for their own research programs. It could be used for a research partnership within a year.
Presented at the 2021 Animal Nutrition Conference of Canada. For information on the next edition, click here.