Sorting Pigs by Weight for all-In/All-Out or Continuous Flow Facilities

By Harold Gonyou and Colin Peterson and published by Prairie Swine Center - Two trials were conducted to determine if pigs should be sorted by weight at the beginning of the growing/finishing phase and if this decision is dependent upon the use of all in/all out (AI/AO) or continuous pig flow management.
calendar icon 26 June 2004
clock icon 7 minute read
Visit the Prairie Swine Centre


Pigs were classified by their relative weight as Heavy, Medium or Light prior to allocation to pens. Within each gender, pigs were allocated to create uniform pens, consisting entirely of Heavy, Medium and Light pigs, respectively, and variable pens consisting of two or more of the weight classes. In addition to standard growth and intake data, behaviour data on aggression and time budgets were collected on the pigs.

There were no deleterious effects of having variable weight pens. Average daily gain (ADG), and behaviour did not differ between pigs in uniform and variable weight pens. The rate of pen and room emptying differed depending upon the pig flow management system and grouping strategy. Under a continuous flow system, pens were emptied in an average of 105.5 days, while under an AI/AO system the rooms were emptied in an average of 107.5 days. Uniform and variable weight pens emptied at the same rate under the continuous flow system. Under the AI/AO system, rooms of variable weight pens emptied faster (104.1 days) than did rooms of uniform weight pens (110.9). Sorting pigs by weight into uniform pens at the beginning of the growing/finishing phase is not advantageous, and may be detrimental to rapid turn-over of rooms under an AI/AO management system.


Pigs ready to be sorted The recommended practice is to sort pigs by weight when they enter the grower/finisher barn. This is due to the belief that creating pens of uniform weights results in greater uniformity at market, earlier emptying of heavy pens, and lighter pigs reaching market weight sooner. However, previous research has not consistently demonstrated such effects, and has suggested that the social behaviour of pigs may be more stable in non-sorted groups.

In continuous flow buildings new pigs are introduced when a pen is emptied, while in all-in/all-out (AI/AO) systems an entire room must be marketed before a new cycle or turn begins. The strategy in a continuos flow system should be to empty pens as soon as possible, while that in AI/AO systems is to get the smallest pigs to market sooner. In continuous flow systems it would seem advantageous to sort pigs by gender and weight so that entire pens reach market weights at the same time and pens can be refilled. However, in AI/AO, sorting by weight may be disadvantageous in that early-marketed pens remain empty while the remaining pigs are crowded into just a few pens. It may be better to place a full range of small to large pigs in each pen, resulting in a social environment with a well defined hierarchy, and pens with increased space for all small pigs once the first pigs are marketed. Thus, the decision to sort or not to sort by weight may be a function of continuous or AI/AO out management.

Materials and Methods

Two trials were run with two replications in each trial. Each replicate included six pens of castrated males and six pens of females, with 12 pigs per pen. Within gender, pigs were classified as heavy (heaviest 33%), medium or light, and randomly assigned within weight class to create three uniform pens, consisting entirely of Heavy, Medium and Light pigs, respectively. Three variable pens were also created, each consisting of two or more of the weight classes.

Pigs were individually weighed on day zero (approximately 25 kg.), then at 2-wk intervals until week 10, and weekly thereafter. Pigs remained in their original pens until they reach market weight, in order to determine emptying time for each treatment. The pen emptying rate, used for continuous flow, was defined as the number of days required for 80% of the pigs in the pen to reach the targeted market weight of 107 kg. The room emptying rate, used for AI/AO, was defined as the number of days for 80% of pigs within a gender to reach the target market weight. Feed intake was determined for each pen on the same schedule as animal weights. Carcass data were collected and evaluated after each trial was completed.

Levels of aggression were observed by live observations on the day of re-grouping to determine the duration and frequency of fighting for each treatment. Observations were also made during the 6th week to determine if weight class and combination method affected the proportion of time spent eating, standing, drinking and lying.

Results and Discussion

Males grew faster than females during each 2-wk period in the trial, with an ADG of 880 vs. 826 gm/day over the initial 12 weeks (P<.01). Heavy pigs outperformed Medium and Light pigs during both of the first two 2-wk periods (P<.01), with 0-4 wk ADG of 794, 747 and 705 gm/day, respectively. However, differences in accumulative ADG among weight classes persisted only until week 10, and no differences existed in the 0-12 wk gains. There were no differences in ADG between grouping strategies (Uniform vs. Variable) during any 2-wk or accumulative period. The relative ADG of Heavy, Medium and Light pigs did not differ between Uniform and Variable pens. Consistent with the results for ADG, males had a greater ADFI than did females (2.52 vs. 2.25 kg/day: P<.01), but Uniform and Variable groupings did not differ (P>.10).

The emptying rates under both AI/AO and Continuous Flow Management were improved for males, compared to females (Table 1), due to their faster growth. Pigs in Uniform pens took longer to empty the room than those in Variable pens under an AI/AO management system, but the difference in average emptying time for pens under Continuous Management was reduced to non-significance (Table 2). Under the AI/AO management system advocated today, random allocation of pigs to pens, within sex, would shorten the length of a turn by more than a week. Carcass data was not affected by group or size.

Pig behaviour was observed and analyzed to study the effects of gender, grouping and size of the pigs on the frequency and duration of aggression. No significant differences in aggressive behaviour were found when comparing the gender, group and size differences in pigs. Behaviour was also observed to assess the effects of gender, group and size on the time budget of the pigs. The proportion of time the pigs spent eating, drinking, lying and standing were compared. There was no difference in the time budgets for size, however male pigs spent more time lying than females, the difference being made up in the time spent standing. Males spent 71.5% of their time lying while females spent 68.0%. The time spent eating and drinking were similar for both genders.

It should be noted that the pigs used in this study would be considered very uniform under most management systems until recently. The pigs were generally within 7 days of age and the weight differences were not large. The results of the study do not support the generally held assumption that sorting pigs by weight is advantageous in terms of subsequent productivity. In fact, when emptying rate is considered, it is advisable to randomly allot pigs, within gender.


Under the management programs of many mid- to large-sized swine farms in western Canada, with weekly weaning and grouping of pigs, there are no advantages to sorting pigs by weight as they enter the growing/finishing barn. Productivity is approximately the same in both sorted and non-sorted pigs, and behaviour responses are similar. This recommendation is particularly relevant to AI/AO systems that are becoming the standard on larger farms.


This research was funded by grants from the Alberta Pork Producers Development Corporation and the Alberta Agricultural Research Institute, under its Matching Grant Program. Appreciation is expressed to Marney Korchinski who assisted with animal work, Stephanie Schmolke who assisted with observations and Kim Getson who analyzed videotapes. Appreciation is also expressed to the numerous other Prairie Swine Centre employees who helped with the project.

Source - Prairie Swine Centre - June 2004
© 2000 - 2023 - Global Ag Media. All Rights Reserved | No part of this site may be reproduced without permission.