Recommendation 2 - Investigate low energy, non-collision events as a source of KF

High energy events within non-cage systems provide a likely mechanism for KF that have been suggested to result from collisions with elements of animal housing such as perches (Scott et al., Reference SCOTT, LAMBE and HITCHCOCK1997; Moinard et al., Reference MOINARD, STATHAM and GREEN2004a; Sandilands et al., Reference SANDILANDS, MOINARD and SPARKS2009; Wilkins et al., Reference WILKINS, MCKINSTRY, AVERY, KNOWLES, BROWN, TARLTON and NICOL2011). Counter-intuitively, KF frequently occur in cage systems where there exists seemingly limited opportunities for dynamic loading as might occur during collision. Thus, improving our understanding of the sources of KF will require insight of dynamic as well as static events. The mechanisms of KF in cage systems are likely not exclusive to this environment, i.e. static mechanisms are likely to cause damage in non-cage systems as well, and thus deserve investigation to provide a more comprehensive understanding of the sources of KBD.

Although commercial laying hens kept in enriched cages and non-cage systems have opportunities for weight-bearing activities, birds’ skeletons remain fragile because of a combination of nutritional, environmental and genetic factors (Whitehead, Reference WHITEHEAD2004a). It would be worthwhile to test whether seemingly innocuous, quasi-static activities can induce KF. While perching, hens place considerable forces on their keels which is in contact with the perch (Pickel et al., Reference PICKEL, SCSHOLZ and SCHRADER2011) and certain perch designs or material may result in severe keel bone deviations and perhaps also KF (Pickel et al., Reference PICKEL, SCSHOLZ and SCHRADER2010). This mechanism would be similar to compression fractures in osteoporotic bones of humans, which have been shown to occur spontaneously or with normally innocuous activities, such as sneezing or twisting (Kondo, Reference KONDO2008). Furthermore, application of modeling techniques, including finite-element models (Tomaszewski et al., Reference TOMASZEWSKI, VERDONSCHOT, BULSTRA and VERKERKE2010) that represent the keel bone, would be useful to provide an understanding of the static and dynamic loading patterns in the bone.

More generally, a broader understanding of the keel, using a comparison of relevant bird phylogenies, could be helpful to establish the morphological capacity of the keel and how modern housing conditions exceed that capacity. It would be useful to assess keel integrity in relation to the functional morphology of diverse avian clades that exhibit significant variation in hindlimb/forelimb modules, keel morphology, flight style, and use of natural perches (Heers and Dial, Reference HEERS and DIAL2012).

Recommendation 3 - Investigate the relationship between deviations and KF

A keel with deviations may lead to unequal bone loading during wing-flapping and concentration of strain energy in ways that increase the risk of fracture. The paired pectoralis muscles are capable of generating enormous force, work and power output during normal flight (Tobalske et al., Reference TOBALSKE, HEDRICK, DIAL and BIEWENER2003) and these outputs increase dramatically during vigorous wing-flapping (Tobalske and Dial, Reference TOBALSKE and DIAL2000; Jackson and Dial, Reference JACKSON and DIAL2011). Comparable wing-flappingis often observed in commercial laying hens that slip from a perch or aviary tier and try to regain their footing (personal observation, M Toscano). Three-dimensional force-balance calculations (Hutchinson et al., Reference HUTCHINSON, ANDERSON, BLEMKER and DELP2005; Baier et al., Reference BAIER, GATESY and JENKINS2006) should be used to test for effects of keel deformity upon bone loading during pectoralis muscle contraction; particularly those observed during episodes where balance is lost or panics (see Recommendation #5). Deviated keels may lead to KF indirectly by complicating balance maneuvers, an additional topic which deserves investigation.

Recommendation 4 - Investigate the role of bird development in KBD susceptibility

It is of prime importance to learn more about development of locomotor and cognitive skills as these contribute towards the ability of hens to navigate within the home system. It is doubtful that the traditional aviary rearing system, where chicks are kept confined to platforms for the first four weeks after which the sides are opened, is the best system to prepare birds for aviary housing (Kozak et al., Reference KOZAK, TOBALSKE, MARTINS, WERBEL and HARLANDER-MATAUSCHEK2015). Under natural conditions, locomotor capacity is critical in Galliformes as they are confronted with immediate challenges to escape predators, search for food and seek shelter (Dial and Jackson, Reference DIAL and JACKSON2011). Beginning at six days post-hatching, wild Galliformes will readily flap their wings to produce aerodynamic forces that enhance hindlimb function while moving up inclines, a behavior called wing-assisted incline running (WAIR) (Dial, Reference DIAL2003; Tobalske and Dial, Reference TOBALSKE and DIAL2007; Dial et al., Reference DIAL, JACKSON and SEGRE2008). Adult Galliformes may prefer WAIR rather than flight to reach an elevated area in a complex, natural habitat (Dial and Jackson, Reference DIAL and JACKSON2011). A better understanding of this process in commercial strains could be helpful to design juvenile and/or adult hen housing systems that improve locomotor abilities (Leblanc et al., Reference LEBLANC, TOBALSKE, WUERBEL and HARLANDER-MATAUSCHEK2015). For instance, variable-engineered systems would be suitable for accommodating the birds as they develop, using more ramps and adjustable angles with increasing age. A potential advantage of WAIR compared with flight may be that whole-body kinetic energy is less during WAIR (Tobalske and Dial, Reference TOBALSKE and DIAL2000; Reference TOBALSKE and DIAL2007), a benefit that could reduce the risk of KBD due to accidental impact with the housing environment. Additionally, increased wing-flapping during development may assist in improving balancing abilities (Filipa et al., Reference FILIPA, BYRNES, PATERNN, MYER and HEWETT2010) while greater activity is known to correlate with bone strength (Rath et al., Reference RATH, HUFF, HUFF and BALOG2000). Beyond musculo-skeletal development, juvenile birds using WAIR will likely learn neuromuscular coordination that will be useful for negotiating three-dimensional structures in adulthood, an impairment suggested by Gunnarsson et al. (Reference GUNNARSSON, YNGVESSON, KEELING and FORKMAN2000). Research in this area will help to identify optimum rearing conditions likely to protect birds from cognitive impairment in adulthood.

Recommendation 5 - Investigate the role of escape reactions as a source of KF

Efforts should be made to understand the damage resulting from sudden escape reactions, or during panic. Escape is normally triggered by a situation that is, or is perceived to be, life-threatening and thus is not ‘normal’ in the context of commercial poultry husbandry, though has been documented (Richards et al., Reference RICHARDS, BROWN, BOOTH, TOSCANO and WILKINS2012). ‘Escapes’ resemble the pattern seen in Galliformes involving take-off using high-frequency, high-amplitude wing beats that feature enormous power output (Tobalske and Dial, Reference TOBALSKE and DIAL2000; Tobalske et al., Reference TOBALSKE, HEDRICK, DIAL and BIEWENER2003), a quick return to the ground using a glide, and then resumption of walking or running. Given that the response is one of last resort, its manifestation is likely to exceed the keel's morphological capacity. Additionally, escape flights will not allow for the precise navigation required in housing systems. A more thorough understanding of the causes of escapes is necessary, including the roles of genetic selection and stockmanship, and the effects of dim lighting conditions and intra-bird spacing (Tillmann, Reference TILLMANN2009).

A comparative evolutionary approach could further aid investigation regarding the role of escape reactions as a source of damage. A museum survey of wild birds of different species showed that 4.5% out of a sample size of 6,212 specimens had sustained and survived bone injuries of which clavicle injuries were the greatest in number, especially in smaller birds, and were attributed to collisions with solid objects (Tiemeier, Reference TIEMEIER1941). The sample included 45 birds in the Phasianidae, the family that includes the chicken, with an incidence rate of 10% in this family. These percentages are far lower than incidence rates of KBD in layer hens, but they do suggest that the escape flight of birds in the Phasianidae (Tobalske and Dial, Reference TOBALSKE and DIAL2000) may be correlated with bone damage. An alternative explanation, however, is that some damage reported by Tiemeier (Reference TIEMEIER1941) was due to gunshot, as many species in the Phasianidae are gamebirds.

Recommendation 6 - Investigate genetic capacity to reduce KBD

The role of genetics as a contributing factor in KBD dates back to work by Hyre (Reference HYRE1955) who showed that the tendency to develop keel deformities was heritable by successfully selecting for and against KBD over six generations. Even earlier, Warren (Reference WARREN1937) showed that crooked keel bones (in comparison to straight keels) had a reduced ash content which he suggested was a causal factor. In considering the scope for genetic selection against KBD, we must first determine what traits should be selected where possibilities include: stronger bones, improved physical ability, and increased docility. Alternatively, as KBD is a relatively complex trait with a number of genetic and environmental factors playing a role, genomic selection should be considered (Fulton, Reference FULTON2012) by carefully monitoring the incidence of KBD in a large population of laying hens and then comparing genomic information of hens with and hens without KBD. One of the strengths of this approach is that it does not target a single factor, e.g. bone strength, but focuses on the actual presence or absence of damage. In theory, the associated mechanism(s) is (are) selected in the process of identifying birds with the desired trait. A challenge with the genomic approach is that it requires a large sampling population (>5,000) of which a clear KBD phenotype is needed. However, once the genomic fingerprint of a hen with no KBD is acquired, no further phenotypic measurements are required for the selection program (Eggen, Reference EGGEN2012). Although breeding for reduced KBD is attractive, one should be aware of possible linkages and trade-offs with other traits, e.g., reduced egg shell thickness and egg breaking strength (Stratmann et al., in press). Whitehead (Reference WHITEHEAD2004b) provides an excellent review of the relationship between skeletal integrity and egg quality.

Recommendation 7 - Investigate housing adaptations that affect frequency of KBD

Large differences between housing systems in the incidence KF indicate that housing design and/or management plays a key-role (Rodenburg et al., Reference RODENBURG, TUYTTENS, DE REU, HERMAN, ZOONS and SONCK2008; Wilkins et al., Reference WILKINS, MCKINSTRY, AVERY, KNOWLES, BROWN, TARLTON and NICOL2011). Perches have received particular attention. As an indication that perches have a causal role in the occurrence of KBD, higher rates were reported at end of lay for hens in conventional cages with (92%) compared to those without (83%) metal perches (Hester et al., Reference HESTER, ENNEKING, HALEY, CHENG, EINSTEIN and RUBIN2013). Similarly, Wilkins et al. (Reference WILKINS, MCKINSTRY, AVERY, KNOWLES, BROWN, TARLTON and NICOL2011) reported a 10-34% increase in KF when perches were added in an organic mobile system. Often, round metal perches are used, which may not offer hens adequate support for their grip. Perches with slightly larger diameters and those made from more flexible materials (wood, rubber) have been suggested to be more capable of absorbing forces during impact and preventing KBD (Pickel et al., Reference PICKEL, SCSHOLZ and SCHRADER2010; Reference PICKEL, SCSHOLZ and SCHRADER2011). Perches covered with a soft rubber layer were successful in reducing the number of keel damage within a commercial aviary suggesting the benefit of this option (Stratmann et al., Reference STRATMANN, TOSCANO, FROHLICH, HARLANDER-MATAUSCHEK and GEBHARDT-HENRICH2015b), possibly not only by reducing the pressure on the keel, but also by providing a cushion and improved grip when landing (Scholz et al., Reference SCHOLZ, KJAER and SRADER2014). Perches as a source of fracture is discussed in more detail by Sandilands et al. (Reference SANDILANDS, MOINARD and SPARKS2009).

Apart from perches, the three-dimensional environment of the hens has to be designed so that it allows the hens to navigate between the different parts of the system. In most commercial aviary systems, improvements are possible that would improve hens’ possibilities for safely navigating through the system. One option is to add ramps to aid transition between tiers. Stratmann et al. (Reference STRATMANN, FROHLICH, GEBHARDT-HENRICH, HARLANDER-MATAUSCHEK, WÜRBEL and TOSCANO2015a) showed that adding ramps to a commercial aviary system reduced falls by 55%, collisions by 41% and keel fractures by 24%, while movements between tiers increased by 44%. The width of the corridors between the different rows within systems also needs further attention. If the corridor width is at the limit of the birds’ navigational ability, it may cause increased collisions due to misjudged jumps (Heerkens et al., Reference HEERKENS, KEMPEN, ZOONS, DELEZIE, RODENBURG, AMPE and TUYTTENS2014).

Another factor to be considered when determining optimum housing is lighting. In many commercial laying hen operations, light intensity is kept at a relatively low level, especially in flocks that are prone to develop feather pecking. Birds need sufficient light and contrast to make an appropriate jump and safe landing (Moinard et al., Reference MOINARD, STATHAM, HASKELL, MCCORQUODALE, JONES and GREEN2004b). The timing of the light transition also seems important: Stratmann et al. (Reference STRATMANN, FROHLICH, WÜRBEL and GEBHARDT-HENRICH2013) reported that vertical movements occurred mainly during a dusk phase when hens usually move to a perching spot for the night. A sudden switch from light to dark periods may increase the risk of KF during this time. Hence, a gradual dawn- and dusk should be investigated as a possible means of reducing KF. Lastly, Heerkens et al. (Reference HEERKENS, KEMPEN, ZOONS, DELEZIE, RODENBURG, AMPE and TUYTTENS2014) found flooring type to also be a factor where wire flooring had greater frequency of KF compared with plastic flooring, though the underlying causes need to be explored.

Recommendation 8 - Investigate nutritional solutions to reduce KBD

Solutions should include changes in bird management and nutrition. The high incidence of KF in cages (Hester et al., Reference HESTER, ENNEKING, HALEY, CHENG, EINSTEIN and RUBIN2013) could point to a calcium shortage. Egg shell formation takes place during the night and hens need a large amount of calcium at this time to produce an eggshell each day. Hens can mobilise this calcium partly from their bones, but they need to restore their supply, otherwise the risk osteoporosis (Whitehead and Fleming, Reference WHITEHEAD and FLEMING2000). One way to supply hens with calcium during the night is to give them daytime access to calcium sources with larger particle sizes, such as grit or shells, which will then be digested during the night. Larger particle sizes of calcium has been shown to benefit skeletal health (Cheng and Coon, Reference CHENG and COON1990; Guinotte et al., Reference GUINOTTE, GAUTRON, NYS and SOURMARMON1995) including that of the keel (Fleming et al., Reference FLEMING, MCCORMACK and WHITEHEAD1998), though usage of this technique varies due to multiple factors including damage to feeding equipment and birds selectively eating the larger particles. Thus, there may be a benefit in supplying grit separately, or developing other sources of calcium that help the hens to restore their supplies and prevent bone weakness. Other nutritional changes could involve incorporation of omega-3 content into the diet which has been shown to result in reduced fracture incidence (Toscano et al., in press; Tarlton et al., Reference TARLTON, WILKINS, TOSCANO, AVERY and KNOTT2013) possibly by modulating bone metabolism and modeling (Liu et al., Reference LIU, VEIT, WILSON and DENBOW2003; Watkins et al., Reference WATKINS, LI, LIPPMAN and FENG2003; Baird et al., Reference BAIRD, EGGETT and FULLMER2008).

Recommendation 9 - Investigate and quantify KBD and production losses

Physical conditions that are associated with pain can, if severe enough, induce redistribution of endogenous resources and derail physiological processes that ensure long-term survival, a classical criterion for compromised welfare (Moberg, Reference MOBERG1985; Broom, Reference BROOM1991). More specifically, Prunier et al. (Reference PRUNIER, MOUNIER, LE NEINDRE, LETERRIER, MORMÈDE, PAULMIER, PRUNET, TERLOUW and GUATTEO2013) advocated changes in productivity as an indication of pain and potentially for compromised welfare. As mentioned before, recent work has shown that individual birds with fractures housed in large groups (~350 birds/group) produced eggs that were characterised with reduced breaking strength and thinner shells (Toscano, in press). This response may represent a diversion of resources where minerals (e.g. calcium) and energy, normally directed towards egg production, must consequently be reallocated to the process of healing bone (Thiruvenkadan et al., Reference THIRUVENKADAN, PANNEERSELVAM and PRABAKARAN2010). Similar results for altered egg production in individual birds were found by Nasr et al. (Reference NASR, MURELL, WILKINGS and NICOL2012b; Reference NASR, MURELL and NICOL2013), although this was assessed in non-commercial conditions (i.e., individual hens isolated in separate cages) in order to link the egg and hen. More critically, work by Nasr et al. (Reference NASR, MURELL, WILKINGS and NICOL2012b; Reference NASR, MURELL and NICOL2013), as well as that by Toscano (in press), did not control for natural variation in bird laying capacity. The lack of pre-KF data leaves open the possibility that birds prone to KF may produce less and weaker eggs independent of whether KFs occurred. Differences in egg character after fracture must be shown to be absent beforehand if the measure is to be a valid indicator of welfare. Interestingly, others have been unable to demonstrate a link between egg production and keel fracture at the flock level (Heerkens et al., Reference HEERKENS, DELEZIE, KEMPEN, ZOONS, RODENBURG and TUYTENS2013). The lack of a relationship could be due to high flock-level variance rather than the absence of an effect, indicating the need for research at bird level responses. Alternatively, Whitehead (Reference WHITEHEAD2004b), reviewing several studies that examined individually housed birds bred for different bone qualities, suggested that little correlation existed between egg production and bone quality. The finding was supported by Gebhardt-Henrich and Fröhlich (2012) who reported more fractures in hens which laid their first egg earlier. As an additional complication in linking the occurrence of fractures with production data, the period in which fractures are seen to most dramatically increase (25 -35 weeks of age) is the one in which birds are coming off peak of lay, thus a drop in production is expected independent of fractures. Therefore, the predicted falloff in egg production resulting from KBD may be subsumed by the drop in egg production as the hen exits the peak of lay period.

Quantification of production endpoints and the loss of productivity associated with KBD can be used as a powerful means to assess changes in animal welfare. More so, because concern for animal welfare is not globally consistent (Lopez, Reference LOPEZ2007) with the strongest interest in Europe and North America, framing the problems of KBD in terms of productivity losses and compromised profit could provide alternative motivations that move towards an ultimate goal of reducing KBD. This particular argument is powerful as it does not diminish the reality that action on the grounds of compromised welfare is necessary, but rather adds a supplementary dimension that will drive stakeholders to effect change.

Interestingly, it is often suggested that KF result from bone that is weakened by the process of demineralisation to provide adequate amounts of calcium for egg shell formation. If correct, continued egg production should associate with a decrease in bone strength and an increase in the occurrence of KF. While this appears to be the case for the first 20 weeks of egg production, recent comparisons of several studies suggest that rates of fractures actually appear to flatten and possibly fall after 45 wks of age (Toscano et al., Reference TOSCANO, BOOTH, WILKINS, AVERY, BROWN, RICHARDS and TARLTON2015; Tarlton et al., Reference TARLTON, WILKINS, TOSCANO, AVERY and KNOTT2013; Stratmann et al., Reference STRATMANN, TOSCANO, FROHLICH, HARLANDER-MATAUSCHEK and GEBHARDT-HENRICH2015b; Petrik et al., Reference PETRIK, GUERIN and WIDOWSKI2015). It is possible that this decrease could be attributed to altered behaviour, though use of an ex vivo impact testing protocol with dead hens (Toscano et al., Reference TOSCANO, WILKINS, MILBURN, THORPE and TARLTON2013) identified a pattern of decreased susceptibility to fracture (Toscano et al., Reference TOSCANO, BOOTH, WILKINS, BROWN, RICHARDS and TARLTON2014) that mirrored the on-farm observations of live hens. Further research is needed to determine how this change in fracture occurrence relates to altered bone physiology and egg production during this period.

Acknowlegements

This manuscript is a product of discussion and collaboration between participants of the 2014 International Keel Bone Workshop held April 2014 in Zollikofen, Switzerland. Organized by the University of Bern, funding was gratefully received from: The Swiss National Science Foundation, The Federal Food Safety and Veterinary Office of Switzerland, the Biological and Biotechnology Sciences Research Council of the United Kingdom, and the Universities Federation for Animal Welfare. Workshop participants, in alphabetical order, were: Michael Appleby (World Animal Protection), Teresa Casey-Trott (University of Guelph, Canada), Rachel Dennis (University of Maryland, USA), Ernst Fröhlich (Federal Food Safety and Veterinary Office, Switzerland), Sabine Gebhardt (University of Bern, Switzerland), Alexandra Harlander (University of Guelph, Canada), Jasper Heerkens (Institute for Agricultural and Fisheries Research, Belgium), Patricia Hester (Purdue University, USA), Darrin Karcher (Michigan State University, USA), Ine Kempen (Proefbedrijf Pluimveehouderij, Belgium), Tine Lentfer (Bundesanstalt für Landwirtschaft und Ernährung, Germany), Maja Makagon (Purdue University, USA), Mohammed Nasr (Zagzig University, Egypt), Prafulla Regmi (Michigan State University, USA), Nadine Ringenberg (University of Bern, Switzerland), Bas Rodenburg (Wageningen University, The Netherlands), Victoria Sandilands (Scottish Rural Cottage, Scotland), Lars Schrader (Friedrich Loeffler Institute, Germany), Markus Schwab (University of Bern, Switzerland), Janice Siegford (Michigan State University, USA), Janja Širovnik (University of Bern, Switzerland), Ariane Stratmann (University of Bern, Switzerland), John Tarlton (University of Bristol, England), Bret Tobalske (University of Montana, USA), Simi Tomar (Central Avian Research Institute, India), Michael Toscano (University of Bern, Switzerland), Stephen Turner (British Egg Industry Council, Great Britain), Teun van de Braak (Hendrix Genetics), Tina Widowski (University of Guelph, Canada), Lindsay Wilkins (Amberly Animal Welfare Consulting, England), and Hanno Würbel (University of Bern, Switzerland).