Data systems for improvement of reindeer production

This paper presents a calf selection program, a recording and a database system developed as a part of a joint Swedish - Norwegian development project where one of the main objectives is to improve identification, handling and production recording systems. The systems are used in a reindeer flock in the south parts of the reindeer area in Sweden. The additional marking system has been used in this flock since 1981 and procedures for identification of the mothers of the calves has been used since 1986.

Expended abstract: The use of the additional marking system in reindeer husbandry as described by Lenvik (1987) and the collar system for the identification of the dams of the calves (Petersson 1990) are essential for recording production data on the individual level.The combination of the individual and owner code is used as an unique identification for each animal in recording systems.This paper presents a calf selection program, a recording and a database system developed as a part of a joint Swedish -Norwegian development project where one of the main objectives is to improve identification, handling and production recording systems.The systems are used in a reindeer flock in the south parts of the reindeer area in Sweden.The additional .markingsystem has been used in this flock since 1981 and procedures for identification of the mothers of the calves has been used since 1986.
During the reindeer production year the following parameters are recorded:

July:
-the owner of the particular calf -the date for observation -the day no. of the year -the day no.within each round up session -the code for weighing equipment -the identity of the calf -the weight of the calf -the sex of the calf -the birth year of the dam of the calf -the identity of the dam of the calf The parameters are recorded in a relation database management system, dbase III (Ashton-Tate, 1984).The field information for the database file, KALVRAPP.DBF, in which the calf registrations are stored is shown in Fig. 1.The output to the reindeer owner from this activity is daily lists, as shown by example in Table 2.In the list, data on the calf, average weight for male and female calf group are given and the caif' s mothers summer results from the last three years.The software behind those lists are dbase III programs.

September:
-the same parameters as above for calves not marked during the summer. September/December/January: -the weight of the calf -the weight of the dam Occasionally, selection of calves for slaughter occurs in September although the main selection activities take place in December and January.For fall and winter selection, a computer selection program for calf slaughter strategies is used where the heaviest calves are kept for breeding.The program has been developed by prof.O. Nissen at the Agriculture University in Norway and is incorporated in the MSTATpackage (Nissen, 1984).reindeer owner gives the figures, as percentage, for the slaughter outtake he is aiming at.The slaughter percentages (sl%), one for each sex group, are transformed in the program to xpoints ( = location on the standard normal distribution) corresponding to one of twenty slaughter percentage classes and stored as a vector.Table 1 shows the sl%, the x-points and the probability enclosed by the standard normal curve above each point representing the probability for selection.
The owner and the identity of the calf are entered into the computer and as soon as the live weight of the calf has been recorded the program recommends «life» or «slaughter» for each particular calf.This recommendation is based on the assumption that the calves weights are normaly distributed and that each calf is compared to a threshold weight.The threshold weight is calculated from the mean, which is continously adjusted, plus the product of standard deviation and corresponding x-point.An example; A female calf's weight is recorded as 42 kg and the sl% for females calves as defined by the calf owner, is 60.The x-point is the 0.25 (Table 1) and the probability for selection is 0.4.During the handling session the average weight for female calves is calculated to 40 kg and the standard deviation is 4 kg.The threshold weight is the computed as 40 + 4*0.25 = 41.00.In this case then recommendation will be to keep this calf for breeding purposes.
Petersson & Danell (1990b) calculated standard deviations, as shown in Table 3, from around 10.000 observations of adjusted calves weights on data collected 1986-1989.As shown the deviation increases with the age of the calves.
Fig. 2 shows a diagram constructed for 5 different standard deviations; 3-, 4-, 5-, 6-and 7 kg-An example; Sl% = 35, mean = 41 and s.d.= 3.From the curve for 3 s.d. and sl% = 35 the threshold weight is derived as 41-2 = 39 kg.That is, all animals weighing more 39 kg are selected.As can be seen from the figure, the size of the standard deviation becomes more important when the fraction slaughtered is either big or small.The weighing/selection program generates a file of production data as shown in Table 4a.Example of the daily output to the individual herder is shown in Table 4b.
After the weighing season the data files from the selection system are transformed to a dbase III file, KALVVIKT.DBF (Fig. 1) and each owner is given a complete list of all his yearlings.Missed calves are specialy indexed.This list is used in other districts outside the owners district and retreived «lost» calved are noted.A herd database: All recordings during the reindeer production year are stored in four different database files which are merged into a database system.The main use of the system has so far been to produce life time production records for the individual female reindeer to be used by the owner as an information in the selection work.
Fig. 3 shows an example of a female life time record that is displayed to the owner during handling of the animals.

Future use of production data:
The information in the database system enables several new areas to be investigated in the field of reindeer researching.For example: 1) Studies of causes of variation in production traits important for choice of selection strategies.
2) Analysis of females life time production ability-3) Analysis of losses.

Table 2 .
Example on daily output during summer handling.