An animal‐based screening method for sufficient amount of straw to fulfil the need for exploration and manipulation
By Margit Bak Jensen and Lene Juul Pedersen, October 19, 2018
This document describes a screening method to assess if pigs are supplied with a sufficient amount of straw to fulfil their need for exploration and manipulation through collection of data on the availability of straw, pigs’ exploratory behaviour and lesion scoring.
• Providing extra enrichment as an early intervention reduced tail biting outbreaks.
• Tail damage was observed among weaner pigs with intact tails in 58 of 60 pens.
• Solitary tail damage did occur without escalating into tail biting outbreaks.
Tail biting is a serious animal welfare problem in the modern pig production. A frequently studied preventive measure is enrichment materials, and increasing levels of enrichment materials, especially litter materials, reduces the risk of tail biting. However, permanent access to litter materials, can cause blockage of the slurry system and increase production cost. The aim of the present study was, therefore, to investigate if providing extra enrichment material, when observing the first tail damage could reduce the prevalence of tail biting outbreaks. The study included 1804 weaner pigs from 7 to 30 kg distributed in 60 pens with intact tails. As basic enrichment material, pens were equipped with two wooden sticks and daily provided with approximately 400 g of fine chopped straw. From outside the pen pigs were checked for tail damages three times weekly. When the first tail damage (fresh or scabbed) was recorded, the pen was assigned to one of four treatments; chopped straw (approximately 200 g/pen) on the floor (straw), haylage in a spherical cage (haylage), hanging rope with a sweet block (rope) or no extra material (control). From first treatment day and until a tail biting outbreak, tails were scored three times weekly. A tail biting outbreak occurred when four pigs in a pen had a tail damage, irrespective of wound freshness. The experiment was designed to compare the prevalence of tail biting outbreaks in each of the extra material group with the control group. A treatment was carried out in 44 of the 60 pens: 10 pens with straw, 8 pens with haylage, 7 pens with rope and 19 control pens. The risk of a tail biting outbreak was significantly lower in pens with haylage and straw compared with control pens (P < 0.05), and there tended to be fewer tail biting outbreaks in rope-pens compared with control pens (P = 0.08). The results should, though, be interpreted with caution due to the relatively small sample size. In control pens with no intervention, a tail biting outbreak developed in 42% of the pens within two to five days after the first tail damage was observed, whereas a tail biting outbreak did not occur in 32% of the control pens. In conclusion, a regular tail inspection and the use of extra enrichment material, when the first minor tail damage occur, could be one way to reduce the prevalence of tail biting outbreaks.
Indirect genetic effects (IGEs) are heritable effects of an individual on phenotypic values of others, and may result from social interactions. We determined the behavioural consequences of selection for IGEs for growth (IGEg) in pigs in a G × E treatment design. Pigs (n = 480) were selected for high versus low IGEg with a contrast of 14 g average daily gain and were housed in either barren or straw-enriched pens (n = 80). High IGEg pigs showed from 8 to 23 weeks age 40 % less aggressive biting (P = 0.006), 27 % less ear biting (P = 0.03), and 40 % less biting on enrichment material (P = 0.005). High IGEg pigs had a lower tail damage score (high 2.0; low 2.2; P = 0.004), and consumed 30 % less jute sacks (P = 0.002). Selection on high IGEg reduced biting behaviours additive to the, generally much larger, effects of straw-bedding (P < 0.01), with no G × E interactions. These results show opportunities to reduce harmful biting behaviours in pigs.
In this study, the possibility of introducing an elevated platform to a piglet pen was explored as a way of increasing available space and creating functional areas. On the platform, nine different manipulable materials were offered. In four batches, 40 weaned piglets were kept for five weeks in the two-level pen. Video recordings were taken two days per week. In the afternoon, more piglets were on the platform than in the morning or at night (7.2 ± 0.1 vs. 4.9 ± 0.1 vs. 0.6 ± 0.1 piglets/5 minutes; p < .05). The area under the platform was preferred more in the morning and at night than in the afternoon (18.5 ± 0.1 vs. 21.6 ± 0.2 vs. 12.5 ± 0.1 piglets/5 minutes; p < .05). Up to 36 piglets were counted there simultaneously, mainly in the recumbent position. On and under the platform, air velocity and ammonia concentration were within the recommended ranges. The study concluded that a two-level pen is a feasible option to increase space allowance and to create functional areas in a piglet pen.
Housing of pigs in barren, stimulus-poor housing conditions may influence their immune status, including antibody
responses to (auto-)antigens, and thus affect immune protection, which will influence the onset and outcome of
infection. In the present study, we investigated the effects of environmental enrichment versus barren housing on the
level of natural (auto-)antibodies (NA(A)b) and their isotypes (IgM and IgG) binding keyhole limpet hemocyanin (KLH),
myelin basic protein (MBP), and phosphorycholine conjugated to bovine serum albumin (PC-BSA) in pigs co-infected
with porcine reproductive and respiratory syndrome virus (PRRSV ) and Actinobacillus pleuropneumoniae (A. pleuro-pneumoniae). Pigs (n= 56) were housed in either barren or enriched pens from birth to 54 days of age. They were infected with PRRSV on 44 days of age, and with A. pleuropneumoniae 8 days later. Blood samples were taken on 7 dif-ferent sampling days. Housing significantly affected the overall serum levels of NA(A)b binding KLH, MBP and PC-BSA, and before infection barren housed pigs had significantly higher levels of NA(A)b than enriched housed pigs, except for KLH-IgM and PC-BSA-IgG. Infection only affected the IgM, but not the IgG isotype. Moreover, changes in MBP-IgM and PC-BSA-IgM following infection were different for enriched and barren housed pigs. These results suggest that the effect of infection on NA(A)b is influenced by housing conditions and that NA(A)b, especially IgM may be affected by infection.
“This guide aims to give practical advice to pig farmers surrounding the complex
issue of providing suitable environmental enrichment to pigs. It provides
useful information from the knowledge of farmers, researchers and scientific
literature on the different ways environmental enrichment can be provided for
differing types of housing and systems. The information is set out in sections
by housing type, and in each, the types of enrichments that are most suited
to each system are discussed, including their properties, how to present
the enrichment, quantities and practical considerations, such as ease of
installation, maintenance and costs.” (cited from the introduction in the guide).
The animal’s emotional state, eventually modulated by environmental conditions, may affect cognitive processes such as interpretation, judgement and decision making behaviour. The Iowa Gambling Task (IGT) is a common method to examine decision making behavior in humans in terms of risk avoidance and risk taking that reflects the underlying emotional state of the subject. In the present study, we investigated the influence of environmental conditions on decision-making in pigs. To assess decision making behavior in pigs, the Pig Gambling task has been developed. In this task, the pig can choose between two alternatives. The pigs can make advantageous or disadvantageous choices, where advantageous, low risk choices deliver smaller, but more frequent rewards, whereas disadvantageous, high risk choices yield larger, but less frequent rewards. In the long run, over a series of successive trials, the advantageous choices will yield more reward and less punishment, where punishment consists of delivering reward into the central food trough, but making it inaccessible. After habituation to testing apparatus and testing methods during the course of approximately 4 weeks, all pigs learned to discriminate between the advantageous and disadvangeous alternatives (acquisition phase) at the age of 9 weeks. After a 14-week retention interval, at the age of 24 weeks, retention performance was tested (retention phase). In both phases, 20 trials per day were given to a total of 120 trials. Saliva and hair samples were collected once at the end of both phases for determining cortisol, and body mass was measured at the end of the retention phase. The pigs increased the number of advantageous choices during the course of training. In in the acquisition phase, barren-housed pigs chose the advantageous options more often compared to environmentally enriched pigs. No differences werer found during the retention phase. All pigs made less advantageous choices at the start of the retention phase than at the end of the acquisition phase. The level of hair cortisol was higher in the barren-housed than in the enriched-housed pigs. This difference was more pronounced after acquisition than after retention testing. No other differences were found for cortisol in saliva and hair. The environments did not differentially affect body mass at the end of the study. Summarizing, housing in a barren environment appears to be more stressful than housing in an enriched environment, as indicated by higher hair cortisol levels in barren-housed pigs, but it also improved acquisition of the PGT.
Tail docking is an undesirable mutilation of pigs. Currently virtually all young piglets are docked in conventional farming so as to prevent tail biting later in life. However, throughout Europe efforts are made to reduce tail docking. Often farmers provide additional enrichment to try and prevent tail biting. Nevertheless, stopping the practice of tail docking may, and frequently does, lead to elevated levels of tail biting, resulting in tail wounds. In relation to this farmers and policy makers would like to know what levels of tail biting would be equivalent to tail docking in terms of pig welfare, i.e. how much tail biting can be allowed before deciding it would be better to continue tail docking. But this poses the problem how to weigh the (lack of) welfare involved in tail biting of a grower or finishing pig against the pain of tail docking of young piglets. Is this possible? And if so, how?
We recently had a brainstorm session on this subject. This is an outline of what we came up with, including a very tentative personal estimate (by MB).
In my personal view when (in the end up to) about 12% of undocked pigs were tail bitten that would be roughly equivalent in welfare to the docking of all piglets. The uncertainty margin, however, is high, at least ranging from 5-25%. The reasoning underlying my estimate is as follows.
Firstly, piglets are normally docked using hot iron cautery. This is quite painful as it involves applying both heat and rather blunt trauma. The heat kills bacteria and thus may reduce the chance of subsequent infection of the tail wound. Tail biting at a later age, by contrast, is caused by even more (and multiple) blunt trauma (due to biting). It also has a substantially higher likelihood of infection. In addition, there is e.g. fear in the tail bitten pig due to being chased by a biter. Based on this I would say that pain (and stress directly related to tail biting) may roughly be about ten times as high in intensity and about ten times as long in duration, compared to tail docking. This would imply that 1 tail-bitten pig is off-set by about 100 docked piglets as regards the intensity and duration of the pain involved.
However, animal welfare encompasses more than just pain. An important additional factor is the level of stress which is not directly related to tail biting activity.
Firstly, there may be stress related to the treatment of tail biting, e.g. when biters and/or victims are taken out of the pen (resulting in social isolation and/or fighting). This stressor, however, is partly offset by the enhanced enrichment normally provided to pigs experiencing an outbreak of tail biting (though not all pigs are equally affected by the ‘costs’ and ‘benefits’). Note that there is another, more macabre, offset involving ‘happiness’ too, and that is the excitement experienced by the (sometimes fanatic) biter pigs when a tail-biting outbreak has started. Note also, that this biter ‘welfare’ is at the same time an indicator of the level of (background) stress experienced by pigs leading to this abnormal behaviour in the first place.
A much more important source of stress that must be taken into account, therefore, is related to the general housing conditions to which the pigs are exposed prior to a tail biting outbreak. Tail biting is an unnatural behaviour that is triggered by (some kind of) stress. Pig farmers are aware of this and will try and prevent tail biting by generally improving the housing conditions when they (start to) raise pigs with intact (undocked) tails. Thus the expected level of stress to which the pigs are exposed is likely to be higher in the case of routine tail docking. When farmers stop tail docking they normally provide much better enrichment (rooting material & space). Farmers raising pigs with intact tails will also take other measures to reduce stress, e.g. provide better climatic conditions, better feed and better health care. These stress-reducing measures don’t just apply to the biters or the victims of tail biting. They apply to all pigs in the pen. Furthermore, they don’t just apply during an outbreak of tail biting, but they apply throughout the pigs’ lives. Hence, the reduced stress levels are a major factor reducing the off-set between docking and tail biting based exclusively on pain (and pain-related fear). I would estimate that the improved living conditions may reduce the off-set by at least a factor 10. This would mean that taking into account both pain and stress, 100(%) docked pigs (kept with minimal care and in a more barren environment) could be roughly equivalent to similarly-sized group of pigs with intact tails under enriched conditions and in which 10% of the pigs has been tail bitten.
Tail biting in docked pigs
However, we know that tail biting does not only occur in undocked pigs. It is also seen in docked pigs. Roughly 2% of docked pigs are tail bitten. It seems safe to assume that the level of pain from being tail bitten is roughly comparable in docked pigs and in undocked pigs (though docked tails may be more sensitive and thus less likely to get bitten). Taking this into account would imply that 100 docked pigs of which 2% also experiences tail biting later in life would be having a level of (poor) welfare comparable to 100 undocked pigs of which 12% gets tail bitten. This is about 6 times as much tail biting as the 2% base-line set under conventional docking conditions.
It must be emphasised again, however, that this level of 12% tail biting is a very rough estimate. So, a wide safety-margin applies, e.g. 5-25%. This may depend in particular on the quality of enrichment and the extra care provided under non-docking conditions.
Please note, that this post is the result of a brainstorm session only and presents a personal view. It illustrates how systematic reasoning (using principles of semantic modelling) can be used to start to answer this rather important welfare question. I have provided a very rough estimate. For a more accurate assessment more detailed studies would certainly be required, both in terms of more carefully including what is already known and in terms of accumulating more empirical knowledge about what is not known yet. At present the assessment is still very speculative, and meant to illustrate primarily how to in principle deal with the question of what level of tail biting is equivalent to a practice of routine tail docking.
Postscript: Excluded aspects and some feedback from readers
Note that, in my estimate I neglected several (minor) aspects.
Firstly, I neglected the fact that for tail docking piglets must be picked up. This results in stress, both in the mother sow and in the piglets. From an evolutionary perspective the procedure of catching piglets may be equivalent to experiencing capture by a predator. This would mean that the given estimate would be a moderate underestimation. However, tail docking may be performed in combination with other treatments such as iron injection and castration. If so, the additional stress from handling may be relatively minor. Note, however, that castration applies only to males and may be banned in the near future, and iron injection may be given orally as a kind of ingestible compost, or as has recently been shown, may not be necessary at all. Hence, combining such treatments with tail docking has a reducing likelihood.
Secondly, I assumed that teeth cutting will not be practiced to treat an outbreak of tail biting, neither in the docked pigs, nor in the undocked pigs. Or, more precisely, at least I assumed teeth cutting is not practiced in substantially different numbers of pig. Such teeth cutting is painful and illegal, so it could be considered appropriate to ignore the practice. However, if it were practiced more in undocked pigs (which are likely to experience higher levels of tail biting), then it would have a substantial impact on the level of equivalence, pushing the percentage back down again substantially.
A third point to note is that I did not include in the estimate other ethical considerations or our (anthropomorphic) emotional responses. An example of the latter may be related to the amount of blood seen in the pen, the farmer’s level of stress (unpredictability) associated to this, and the potentially adverse economic consequences associated with tail biting. An example of other ethical considerations is the fact that tail docking may be considered to be an infringement of the animals’ integrity or intrinsic value. In such a rights-based moral view tail docking may be considered ethically wrong, regardless of the level of tail biting when tail docking is stopped. Such aspects were excluded because these are aspects not directly related to animal welfare. They are more related to our human perception of ethics and/or human welfare, rather than animal welfare.
Finally, it is most important to emphasise that I have considered steady-state conditions, but realize that all practices are subject to optimisation. The practice of tail docking has already been optimised for over a period of at least 50 year. By contrast, the practice of raising pigs with intact tails still more or less has to enter the phase of optimisation in commercial practice. This implies that substantially higher levels of tail biting may be regarded as acceptable, provided this is only temporary and provided it leads to substantially lower levels of tail biting later on. In other words, it requires that farmers will persist in raising pigs with intact tails and have a chance to learn to deal with it over a certain transition period, both in terms of prevention and treatment of tail-biting outbreaks.
Feedback reader 1:
Regarding the painfulness of tail biting vs tail docking, I find it impossible to guess the relation – especially as tail biting comes in so many forms.
I absolutely agree that a weighing like this is necessary, but I also think it is a bit dangerous to throw out estimates that are not really based on any evidence (or at least you do not present any?), such as the 100 times worse pain experienced by bitten pigs than docked pigs. Also, tail biting is very heterogeneous, from just a small, one-time bite, to a chronic situation, where the entire tail is lost, so the way you estimate the pain simplifies the matter greatly.
As to the expected level of actual tail biting when docking is stopped: I estimate a two-fold increase in tail biting if no docking is performed. Perhaps somewhere between 2- and 4-fold, based on e.g. slaughterhouse data. There may be a 4-fold increase when the housing situation is not improved otherwise – which you also take into account in your text – when applying a non-docking policy the farmer would normally also improve housing conditions, thus reducing the risk further. I certainly agree that when a farm stops docking, they will probably have a higher incidence of tail biting initially, but on the long-term (as is shown e.g. in Finland where tail docking is totally forbidden, and the tail-biting incidence, based on abattoir data is around 2%), a 10 or 12% incidence is certainly higher than I would expect.
Feedback reader 2:
Having read your blog I think you need to factor in adaptive, compensatory pain modulation into your model.
It is sometimes too easy to fail to take into account post-injury peripheral and central modulation of pain signalling that occur as part of the normal healing process and only focus on the ‘pro-pain’ component.
I also don’t see how you can substantiate this claim?
‘Based on this I would say that the pain of tail biting may be about ten times as high and about ten times as long, compared to tail docking. This would imply that 1 tail-bitten pig is off-set by about 100 docked piglets as regards the intensity and duration of the pain involved’.
While I think it might be possible to attribute weighting to some risk factors within systems, I don’t think it can be applied to pain experienced by an individual (or even at group level as you are suggesting) because there are so many factors that contribute to an individual’s experience of pain? I don’t think you can quantify the painfulness of tail biting and tail docking.
Also when thinking about stress you might want to define what you mean by that in relation to chronicity?
Short-term compensatory responses to stress are in my view positive for the animal; however beyond that when there is a failure of compensation and ultimately homeostatic decompensation then they are undoubtedly negative.
I guess I’m suggesting that any weighting approach might need to accommodate (or factor in) changes over time (i.e. dynamic weighting?)
I hope you find my comments helpful?
As to substantiation, again, it’s my suggestion for a start of an argument to answer this in my view fairly important question. My answer is based on my personal experience as a vet and scientist, and on reasons indicated in the blog. It is certainly in need of further study, examination and assessment. I fully acknowledge the considerable level of uncertainty as well as the risk associated with trying to answer the question. At the same time, however, I would also argue that there is a considerable risk in refusing to try to answer the question, as this leaves the issue to stakeholders.
Feedback reader 3:
Joining the discussion rather late, but basically I agree with the points others have made. I think it quite reasonable to conceptually set out the trade-offs which would determine the level of tail biting above which tail docking could be ethically justified, but putting numbers on some of these things is rather difficult.
For risk of tail biting in docked and undocked pigs we have a growing number of published sources and comparative national data.
For experimental comparisons we have old data suggesting increases of 30-60% in pigs in unbedded systems.
More recently we have studies suggesting somewhat lower results if straw is given.
So this part is perhaps simple, but depends on your assumptions about which husbandry systems will pertain across Europe.
For the welfare detriment of tail docking and tail biting, data indicate that both have long lasting effects on pain processing pathways, but the implications of this for pain perception for the individual are uncertain.
For tail docking, the data I have seen are still contradictory on whether cautery is more or less painful than simple section (some suggest the cautery destroys the nerves whilst others suggest greater pain). There is also the possibility of tail docking with anaesthesia/analgesia as a route of adoption.
For tail biting, the short term pain will certainly depend on the severity and, even more, on the prevalence of infection. The data on this are currently lacking to my knowledge.
The welfare impairment of keeping in conditions which give rise to tail biting is clearly the greatest of all in magnitude (severity x duration x no of animals) but I don’t think we have any way of comparing the welfare severity of ‘behavioural frustration’ against that of injury/pain. I would be concerned about taking arbitrary figures in the absence of any logical basis.
So, I guess my suggestion would be to explore the framework for this decision, but be very wary about pretending we can quantify it.
I also think the issue not addressed in your blog is the time course of any transition to cessation of tail docking and how to manage this. What proportion of farmers would have the awareness, capital and staff training to implement the changes necessary to their existing housing if obliged to cease tail docking (some older, fully slatted and large group housing systems will pose much bigger challenges and possibly require replacement of buildings), and how long would it take across Europe to reach the ‘acceptable’ situation of relatively low differential in tail-biting prevalence between docked and intact tails, rather than the ‘unacceptable’ differential shown for “one off” change in tail-docking experiments (stopping docking without further improvement of the environmental conditions). I think it important to highlight that your analysis relates to a ‘steady state’ situation and the importance of how any transition is managed and the welfare implications which this will have.
Note that I have not been comparing docking versus non-docking in a mono-factorial way. I compared docking in a more barren environment versus not docking in a more enriched environment supplemented with special attention by the farmer, as that is what will normally happen in practice. I have now emphasised this more clearly in the text.
I largely agree that we currently largely lack the data needed to quantify more precisely. However, I also believe that in principle it is possible to do so, and that the estimate/assessment can be more or less verified empirically (as the body of knowledge accumulates and modelling principles are improved). Personally, I am inclined to try and quantify despite considerable uncertainty, because it provides a better starting point for further discussion. In addition, such preliminary but more science-based estimates are much needed to complement the inevitably politically-loaded figures and personal assessments presented by farmer-representatives and NGO’s arguing either (rather exclusively) against or in favour of ending tail docking as a routine practice to prevent tail biting.
An important point I’ve been trying to make is that pain is not the only relevant aspect of welfare involved in tail docking and tail biting, and that the levels of enrichment and care should also be taken into account. I don’t think it is even possible to honestly say it is not possible to ‘add’ these aspects, since proper political decision making (in all kinds of areas, not just tail biting) simply does and has to, whether it is considered scientifically possible or not. And if so, I would argue it is most reasonable to try and provide the best possible scientific support, while being as honest as possible e.g. about uncertainty margins and the relevance of incorporating more information. I also think the estimate provides broad support to ‘farewell-dock’ initiatives such as those in Finland, Sweden, Denmark, the Netherlands, the UK and Germany.
• Sows made contact with rope more frequently than rubber sticks or fixed woodblock.
• At any given time, more sows made contact with rope than the rubber or woodblock.
• Rope preference observed during day and night for two weeks.
• No difference in the lesion severity among the treatment pens.
Increased public concern about farm animal welfare is driving both legislative initiatives and market forces to change how sows are housed and managed. This study investigated the use and preference for enrichment items at a 5600 sows commercial sow farm in eastern USA. Gestating sows were housed in static, pre-implantation groups of approximately 75 sows per pen and fed via a single electronic sow feeding station. Each pen contained one of three enrichment objects (OBJ): hanging rope, hanging rubber sticks, and a fixed wood block. Behavioral data was collected from 18 pens during the course of this study on days 1, 3, 5 and 14 (DAY) that sows were in the pen, and at specific times on each day (TIME). For daytime activity, data was collected on-site in three 2-h blocks between 0800 and 1000, 1100–1300 and 1400–1600 for each pen and for nighttime data was collected on Day 1 in three 1-h blocks between 2200 and 2300, 0000-0100 and 0200-0300. Behaviors recorded included proportion of observation time animals interacted with the object, proportion of animals in pen that interacted with the object, and posture (up/down) of each animal in the pen. Lesion scores were recorded prior to mixing and two weeks post-mixing as a proxy for social aggression. The median proportion of observation time that the sows were in contact with the rope (62.4%) was significantly greater than (P < 0.01) the median proportion observed in the rubber pens (31.5%) and significantly greater than (P < 0.01) the median proportion observed in the woodblock pens (24.3%). Mixed design ANOVAs indicated a significant interaction of OBJ and DAY (P < 0.01) and OBJ and TIME (P < 0.01) on the proportion of observation time that the sows were in contact with the enrichment objects. Post-hoc analyses using Bonferroni correction showed that on each observation day and time period, the proportion of observation time that the sows were in contact with the enrichment was significantly greater (P < 0.01) in rope pens than rubber or woodblock pens. These results indicate that sows can exhibit clear preferences for enrichment type, with the sows interacting with the rope significantly more often throughout the study, at each sampling hour. However, there were no significant differences in lesion severity or sow activity between the three enrichment types, suggesting that common behavioral patterns including the establishment of social hierarchy took precedence over the pursuit of available enrichment. Additional studies are needed to understand how preferences for enrichment objects could be utilized to potentially impact sow productivity and welfare.