Review of the potential impacts of freight rail corridors on livestock welfare and production

Literature search protocol

Initially, generic search engines were used to source news articles, opinions and other resources to help inform and direct the literature search terms to include public and producer concerns regarding the impact of rail infrastructure development. Then, the following protocol was utilised to identify relevant literature: first, a search was conducted investigating the impact of rail corridors on domestic livestock, and then the criteria were broadened from there on the basis of particular focus areas. Electronic databases (Google Scholar) and the University of New England library catalogue integrated with the Primo library discovery service were used, drawing from several scientific databases such as ProQuest, EBSCO, Scopus, and Web of Science. Additionally, news articles, opinions and other resources, including public and producer concerns regarding the impact of rail infrastructure development, were sourced. Search terms included, but were not limited to, livestock, rail corridors, noise impacts, anthropogenic disturbance, animal production, train, and livestock underpass.

Literature screening

Literature was assessed on a number of inclusion criteria. When assessing literature pertaining to wildlife, these were restricted primarily to studies on impact of rail and road traffic, as well as underpasses and wildlife corridors when looking at impacts on predation behaviours. Given the lack of specific research in rail impacts on domestic livestock, it was necessary to broaden search limits to allow the identification of literature describing a range of impacts and interpret how they might give insight in this context. Literature examining the impacts of noise on animals specifically focused on livestock in normal production environments such as paddocks and yards.

Research looking at the impacts of rail on wildlife biodiversity and the environmental impacts of rail corridors were excluded, except where this may affect the behaviour and welfare of livestock.

Searches were conducted using the common names and scientific names of the livestock species of interest (primarily sheep and cattle) and the terms ‘rail’, ‘anthropogenic disturbance’, ‘noise’, ‘impact’, ‘train’, ‘corridor’, ‘noise stress’, ‘sound’ ‘underpass’, ‘effects of noise’, ‘auditory’, ‘visual’, by using various combinations to broaden the search base. As well as searching literature published in English, we also encountered resources during our search that required some translation, which was performed using the ‘Google translate’ function.

Data included in Table 1 were assessed and included conditional on meeting the criteria of relevance to potential impact of rail noise and included details on species, noise level, and behavioural or production impact measures and results. Literature was selected for inclusion on the basis of relevance to each focus area, either in terms of impact on the animal (physiology, stress etc.) or the stimulus itself (noise, vibration etc.), by using the following inclusion-topic criteria:

The purpose of this review was to investigate the potential impacts of a freight rail corridor on neighbouring livestock production enterprises, with a (Sibbald et al. 2009) focus on impacts on the animals themselves and how noise, vibration and visual disturbance may affect the way in which they live their lives and respond to threats. This review does not aim to make animal welfare assessments or estimate production impacts.

Authors chose to work within focus areas that aligned with their fields of expertise to be as comprehensive as possible within each context of animal production. These fields encompassed animal welfare, animal health, stress physiology, animal behaviour, reproductive physiology and behaviour, grazing and spatial behaviour, predation, production, animal handling and farm practices. For this reason, screening was completed for each section separately, by one author per section. A framework was proposed utilising the data in the literature to attempt to place rail disturbance within the context of known stressors. Due to the lack of research on direct impacts of rail proximity to livestock, the framework was based on potential and likely impacts of the isolated and cumulative effects of stressors associated with rail corridors.

Personal communications were sought from fellow researchers who have worked with livestock within proximity to rail lines for their anecdotal observations to validate conclusions from the review of literature, particularly when findings from species other than domesticated ruminants were extended to these species.

Auditory (noise) and vibration disruption

Livestock in general, being prey animals, are sensitive to sounds and tend to respond to sounds by orientating their ears towards the sound source. Studies on livestock sensitivity to sound have predominantly focused on sound intensity (loudness) more than sound frequency (pitch). Pigs, sheep and dogs have similar auditory ranges (Heffner 1998), and may have similar aversion to high noise environments. Sheep and cattle have the ability to hear sounds in the ultrasound range (above 20 kHz), a range which is inaudible to humans (Adamczyk et al. 2015). Sounds in this range, such as rail wheel squeaks, may affect how the sheep respond to machinery (Fourie 2018). Previously, Brouček (2014) investigated the impact of noise across a range of sources, although not focusing on noise associated with rail corridors.

Noise can mask, or inhibit, the perception of sounds (Barber et al. 2010) and has the potential to affect behaviour and communication between animals. This masking is not just related to hearing acuity, but also the cognitive perception of auditory stimuli (Gutschalk et al. 2008). While local background ambient farm noise levels may be difficult to quantify, previous measurements by Thorne (2010) estimated ambient levels at a study farm to be in the range of 34 dBA and 35 dBA. Other characteristics of the sound (continuous or intermittent) contribute to its stressfulness. Sheep have shown ability to adapt to continuous noise, whereas loud, unpredictable, intermittent sounds are often recorded as the most aversive to sheep (Weeks 2008). Sheep show evidence of an ability to habituate, or adjust, to 100 dB intermittent noises (including electrical and diesel engines, jet and propeller planes) within 9 days (Ames and Arehart 1972), while cattle habituated to a recording of abattoir sounds more rapidly than the visual and physical activity associated with it (Grandin 1997). Further to this, Haas and Scrivener (2015) described the impact of high-speed rail on equines in terms of rapid onset rate, i.e. how quickly a loud noise is produced and is measured in decibels per second, resulting in a potential startle effect on animals. The recorded effects of noise on several species of domestic animals is described in Table 1 and demonstrate the knowledge gaps associated in the area of noise and livestock, particularly around habituation. Noises at almost three times that of ambient background farm noise generally result in responses such as increased movement or vigilance behaviours, and increased heart rate, but animals also show capacity for habituation, often within several days (Waynert et al. 1999; van Der Staay et al. 2011). While the impact of high-level (120 dB) noise is known to cause hearing loss in sheep and other animals (Griffiths et al. 1994), more subtle effects of lower-level noise such as masking of communication (Rosa and Koper 2018) are less well known.

Noise as a stressor has been shown to reduce the quality of farm animals’ life, with sounds of approximately 100 dB emerging from the literature as a possible threshold for stress (De la Fuente et al. 2007; Voslarova et al. 2011; Kim et al. 2014). Evidence for thresholds of noise having an impact on domestic ruminants (sheep, cattle and goats) is limited. Below 100 dB, sheep and cattle tend to respond behaviourally; movements such as startle reflexes, or moving away from the sound, show a level of aversiveness, and behavioural responses are a first-line defence against greater impacts on animal welfare at a physical or physiological level. However, even at volumes of ~80–100 dB, sheep and cattle show an ability to habituate to noise, reducing their behavioural responses, often within 5–10 days (Rylander et al. 1974; Waynert et al. 1999).

Exposure of domestic pigs to repeated noise stress cause changes in neuroendocrine regulations, which are characterised by temporal alterations in the responsiveness of the hypothalamic–pituitary–adrenal (HPA) system (Kanitz et al. 2005). That research concluded that repeated exposure of pigs to noise levels of 90 dB affects HPA function, resulting in a state of chronic stress that may have negative implications on animal productivity and welfare. Pigs exposed to 90 dB prolonged or intermittent noise had increased cortisol, and noradrenaline to adrenaline ratio (Otten et al. 2004). Pigs are very sensitive to noise and should not be exposed to constant or sudden noise. Therefore, noise levels above 85 dB must be avoided in buildings where pigs are kept (Fottrell 2009). Chloupek et al. (2009) determined a significant negative influence of noise exposure (80 and 100 dB) on the stress and fearfulness of broiler chickens, while Campo et al. (2005) exposed laying hens to 90 dB truck, train and aircraft noise for 60 min, finding increased stress and fear responses.

Distance from the source of the noise can also ameliorate many of the negative behavioural effects. For example, sheep exposed to the sound of a machine gun showed reduced behavioural signs of disturbance the further from the source they were, and no behavioural signs were observed at a distance of 328 m from the source of the noise (Hauser and Wechsler 2013). Because behavioural responses to stimuli are the primary reaction of animals, livestock will first seek to move away from any rail stimuli they find aversive, and once the stimulus has decreased (as the train passes), may return. Thus, impact of noise and vibration is likely to be greatest where livestock are detained in small areas, such as in handling yards, or in intensive grazing cells, and lowest where paddock dimensions allow stock to move away from the rail line. The impact of distance from the rail line can also be utilised in management strategies, such that stock that may be most vulnerable to the effects of rail stimuli, such as young lambs and new stock who have not been previously exposed to the rail line or ewes in late gestation, are located in paddocks at a greater distance from the tracks.

Noise created by the train will occur congruently with the vibration as the train approaches, and then passes by the livestock. It is therefore difficult to separate the two potential stressors from each other. Previous studies have reported the effect of vertical vibration in poultry, rodents and livestock during road transport (Randall et al. 1997; Abeyesinghe et al. 2001; Garcia et al. 2009; Doggett 2018). However, no known studies have assessed rail vibration on livestock productivity or welfare. It is likely that the intensity and duration of vibration experienced by animals being transported is far greater than that experienced by animals in a paddock near a railway, because the extent of animal behaviour and physiological effects rely on the frequency and magnitude of the vibration (Rabey et al. 2015). The vertical vibration levels experienced by animals being transported are constant for the duration of the trip and have been reported to vary from 0.5 Hz to 25 Hz, whereas railway-derived vibrations are generated primarily at the wheel-rail interface, ranging from 2 to 80 Hz depending on track structure and train velocity (Fahy and Thompson 2015), and then travel into the soil (i.e. ground–borne vibrations). One important aspect to be considered is that the ground-borne vibrations propagating from the track can be felt up to 50 m away (Sheng et al. 1999), which is likely to be below the threshold of effect in livestock.

Ground-borne vibrations also result in ground-borne noise; in that case, rail traffic at night-time brings vibration, noise and light to the landscape, possibly causing stress by disrupting sleep in nearby sheep. The effect of freight train noise and vibration on sleep in humans has been investigated (Croy et al. 2013), with sleep fragmentation (Smith et al. 2016) and increased heart rate during sleep having been observed. While the impact of rail noise and vibration on animal sleep is not known, chronic disturbed sleep can have long-term impacts on neurodegeneration (Owen and Veasey 2020). Sheep have shown sleep characteristics similar to those in humans (Schneider et al. 2020) and similar effects of sleep deprivation (Perentos et al. 2016). Interactions between sleep and noise have been studied in laboratory animals, with Rabat (2008) reporting environmental noise resulting in more sleep disturbance in rats than from white noise. Stress responses are activated by noise through the autonomic nervous system, with levels even below accepted thresholds showing impacts (Turner et al. 2005) and intermittency of noise having a greater impact (Rabat et al. 2004). Ground covering or vegetation, including groundcover, may help reduce or diffuse the direct travel of rail vibration through the paddocks or pastures in which the sheep are housed (Fahy and Thompson 2015). Offering livestock the option to move/rest away from the railway by avoiding the use of small paddocks in the proximity of the railway can give the animals the opportunity to choose whether or not to experience the negative stimulus (Kearton et al. 2020).

Visual disruption

There is some evidence that moving trains, carriages or vehicles per se are an aversive visual stimulus for domestic livestock (H. E. Salvin, pers. comm.). There is no evidence that particular colours or materials may be more aversive than others to livestock and nor is there any evidence that the increased speed of movement of the perceived threat, object or vehicle is associated with increased stress. At close proximity, in handling yards, moving vehicles can be a distraction to livestock, and cause them to baulk. In combination with the auditory and vibration effects when stock are housed close to the rail line, there could also be a tendency for sheep to exhibit startle or escape behaviours and flee when the train is in sight.

Designing yards and races with blocked-out lines of sight of moving vehicles is a long-standing recommendation. Location of yards close to rail corridors may cause problems with ease of handling of sheep, unless they are habituated to passing rail traffic. Tree lines or vegetation may be utilised to break up the visual impact of the trains as they pass by, creating a ‘safe’ barrier between the perceived threat and the animal. Over time, habituation to rail traffic will reduce the perceived threat of moving trains, and the accompanying stress.

Physical rail infrastructure

The physical infrastructure of the rail corridor consists of the railway line (fenced off) and level crossing or underpasses for stock passage across the railway line. The subdivision of paddocks by the rail corridor will reduce paddock size in a fashion similar to a fence line. However, where a simple fence line provides only a physical and visual barrier to free livestock movement across the landscape, a railway line has additional impact in the aversive nature of the rail stimulus (outlined above), but the impacts are not specific to the landscape division, and so are not addressed in greater detail here. Stock underpasses are a unique characteristic of a railway line, and potentially offer positive and negative impacts to livestock production systems.

Underpass and overpass use by wildlife has been extensively investigated, with a number of known and potential impacts. Impacts of use by domestic animals depends on factors such as drainage, maintenance, and whether the underpass will be designed for animals to travel freely across areas divided by the railway or whether it is only for controlled movement (Casburn and Cumming 2009).

While there is potential for predators to exploit the use of an underpass by prey animals, there is little evidence for this occurring (Little et al. 2002; Martinig et al. 2020). Underpasses could provide shade and shelter with the correct design enabling free access, and enables safe moving of livestock from one side of the track to the other (James 2020).

To mitigate the risk of predators congregating around underpasses, it is recommended that monitoring and control measures incorporate stock underpasses as a potential predation-risk area. If the animals have free access to the underpass, it is important to ensure that there is adequate drainage so that the animals are not traversing persistent stagnant or muddy conditions, which may lead to foot rot and other bacterial infections (Abbott and Egerton 2003). Design of the underpass should also consider options that facilitate the flow of movement through the area, avoiding dark openings, uneven footing, and sharp corners (James 2020). Inclusion of a light source in the underpass itself, such as, for example solar-powered lighting above animal passageway, will encourage animals to move effectively from entrance to exit. Vegetation around the entrance can reduce noise and visual disturbances around the underpass, facilitating stock movement (Casburn and Cumming 2009).

Behavioural responses to rail corridor stimuli

The combined stimuli resulting from the construction and on-going traffic of a rail corridor may elicit a range of behavioural and physiological effects on livestock, which can change over time. The characteristics of each specific stimulus need to be assessed in terms of their predictability by the livestock. If a certain level of predictability can be established or ‘learned’ by the animal, then a process of habituation can occur, whereby the stress that a particular stimulus elicits in the animal is reduced, either by a change in how stressful the animal perceives the stimulus to be, or by coping mechanisms implemented by the livestock in response to the stimulus. Stress responses and coping mechanisms are manifested in behavioural responses to the various rail stimuli. If the animal does not reduce its perception of the stressfulness of the stimulus or is not able to implement effective behavioural coping mechanisms, then physiological changes may occur. Temperament is an important consideration when investigating the impact of a novel stressor, and reactions can vary between and within species and breeds within species (Lanier et al. 2000; Atkinson et al. 2022).

Predictability of rail stimuli

Predictability of a negative stimulus is known to influence the subsequent stress and behavioural responses in animals (Weiss 1972), with lambs showing a reduced startle response to a sudden event that was signalled with a light cue compared with those that had received an un-signalled event (Greiveldinger et al. 2007). Furthermore, sheep that are able to control their negative experiences have shown reduced stress responses in virtual-fencing case studies (Kearton et al. 2020).

Although the timing of trains may not always be predictable or consistent, a passing train could be considered a predictable stimulus due to (1) considerable auditory and visual signalling as the train approaches, and (2) consistency of ‘behaviour’ from the train.

The ability of an animal to determine the speed of an object via temporal changes has been observed to have a relationship to body size and metabolism, with smaller animals with high metabolisms showing an ability to perceive changes over finer timescales than do larger animals (Healy et al. 2013). This ability may also be confounded in the case of vehicles and trains, which can move much more quickly than does a natural predator, meaning that timing of the sound and visual approach may be unable to be effectively processed by the sheep’s sensory and neurological capacities. Therefore, sheep and cattle may be less able than are smaller animals to accurately predict when an approaching train will arrive, possibly explaining observations of persistent startle and avoidance behaviours even when these species are accustomed to passing trains. However, due to the lack of literature available, it is difficult to determine whether this sensory capability affects the ability of sheep or cattle to effectively acclimate to passing trains.

During the construction phase of the project, a variety of human and machine activity will result in low predictability of stimuli, but sheep and cattle will quickly learn the constrained spatial impact of the activity. Once the phase of rail traffic adaptation begins, the consistency of train ‘behaviour’ will increase the speed at which sheep learn the limitations of train movement and the potential impact. This is essential for the process of habituation to occur, to minimise the duration of this phase, and for stock to move into the final phase of normalised rail traffic.

Acclimation to the new rail traffic and environment may take some time to occur, with some work finding that sheep adjusting to a new environment had not fully acclimatised to their new surroundings within a 9-week study period (Fordham et al. 1991). The low predictability of activity during construction phase may result in sheep avoiding the area close to the activity during this phase. High predictability during the adaptation to rail traffic phase will shorten this phase, but to reduce the incidence of a startle response in sheep, and promote neutral interactions with rail traffic, grazing sheep within sight and sound of rail traffic, but not containing them close to the track, is recommended in the initial weeks of Phases 1 and 2. Predictability does not only rely on a temporal pattern of rail traffic, but also on low-impact signalling of a more stressful stimulus, such as, for example, an increasing volume of noise from an approaching train before it passes close to stock. There is insufficient information in the literature to make a firm recommendation on a distance at which a startle response is less likely to occur, but allowing stock freedom to respond behaviourally, for example, with flight, will aid in the process of habituation.

Habituation

Habituation occurs when animals are exposed to the same stimulus repeatedly, and eventually stop responding to that stimulus. Suddenness, unpredictability and unfamiliarity combined serve to enhance physiological (e.g. heart rate) and behavioural (e.g. startle or flight) responses to a stimulus (e.g. a passing train), while familiarity can serve to modulate the behavioural response to a sudden event (Désiré et al. 2006). Sheep have also shown reduced responses over subsequent exposures to sonic booms, suggesting that sheep can become accustomed to loud noises, although some sheep still showed avoidance behaviours even after repeated exposure (Rylander et al. 1974). Avoidance behaviour was also observed in dairy cattle exposed to noise, with 81% of heifers choosing to avoid noise at 85 dB when offered the choice (Arnold et al. 2008). A strong habituation effect was observed in work by Rylander et al. (1974), investigating behavioural responses to sonic boom and subsonic flight. Cattle showed slight startle responses to the boom effect, returning to neutral within 10 s. Sheep showed reduced reactions to flyovers over time but were not exposed to sonic boom, so could not be compared to cattle. Ducks showed interrupted behaviour on exposure to the boom effect; however, avian responses varied greatly (Rylander et al. 1974). These differences are highlighted in the findings of Trigg et al. (2023), who found that avian species were affected at lower noise and vibration levels than were mammalian species.

Such physiological and behavioural responses can be energetically expensive, and so through the learning process, habituation allows animals to refine their assessment of the threat a stimulus, such as a passing train, poses, and only respond as necessary. The speed of habituation can be highly variable, depending on the temperament of the animals, the frequency of exposure to the stimulus, the predictability of the stimulus, and the learned consequences of the experiences the animals have with the stimulus.

Additionally, risk-disturbance theory may play a role here. This theory suggests that some areas may be avoided due to human disturbance, but can be regulated by the attractiveness of the resources (Lowrey and Longshore 2017). Therefore, while grazing animals may initially avoid grazing close to human activity, if the area is comparatively attractive such as a higher quality of grazing, they may show a higher tolerance to this activity.

Generational habituation (Guinn 2013) is also a potential benefit that may be utilised in some contexts, and has been observed in wild animals habituating to the proximity of humans (Schell et al. 2018). Generational learning also occurs in domestic livestock, the practice known as ‘hefting’ in some regions in which flocks remain in a home range (Davies et al. 2008), or the practice of selling properties with stock in place where generational knowledge may be preserved. Whether generational learning can occur in livestock in the context of noise and visual disturbances is less well understood.

Throughout this review, there is acknowledgement that habituation may occur and therefore mitigation strategies that may reduce impact during the initial exposure to rail stimuli will potentially become obsolete over time.

Behavioural responses of individuals

A behavioural response to a stimulus is the animal’s first line of defence against stress. A well-understood example is a flight-behaviour response to a perceived threat. Other changes in behaviour that could be expected from rail stimuli (at least initially, and potentially long term) include increased vigilance (visually checking for and assessing potential threats), startle and flight behaviours, changed use of the landscape close to the railway line, among others. These are discussed in detail later in this review. A change in animal behaviour alone, from rail stimuli, is not an indication of increased stress per se, but may be an animal seeking to avoid a potential negative impact from that stimuli. However, changes in behaviour can have negative impacts, for example, if increased vigilance reduces grazing time. Alerting distance can also vary across species and environments to novel stimuli; for example, polar bears detected the approach of a snowmobile over 5 km away, but avoidance responses varied significantly (Andersen and Aars 2008). Behavioural responses to rail stimuli will change over time as animals become habituated to the stimuli associated with it, learning about how trains ‘behave’ and refine their assessment of the threats they pose.

Behavioural responses to the introduction of a rail line in their environment will vary depending on the stage of development, human activity in the area, and time for habituation to occur. It is well known that many wildlife species change their behaviour in response to human-altered landscapes, but it is less evident how this may affect livestock, given their management already involves a certain degree of human disturbance.

Livestock that are habituated to handling show reduced stress responses (Tamioso et al. 2017; De Palo et al. 2018). Habituation results in a reduction of the magnitude of hypothalamic–pituitary–adrenal (HPA) axis activation; however, the degree of this response can be affected by other factors, such as predictability, in which the more predictable the stimulus the more rapidly an animal will habituate. An animal may become habituated to other stimuli in addition to the original stimulus, known as generalisation (Grissom and Bhatnagar 2009). This generalisation may have positive implications for production, such as reduced reactivity to novel stressors, conversely there may be a reduction in predator awareness and vigilance behaviours.

The behaviour of ruminants other than domestic sheep or cattle, such as wild elk and pronghorn, in response to road traffic noise has shown reduced responsiveness (vigilance, flight, travelling and defensive behaviours) with an increasing vehicle traffic (Brown et al. 2012). It was suggested by Brown et al. (2012) that these animals could not afford to maintain responsiveness to the most frequent stimuli, and while this might be beneficial in terms of energy investment, detection of predator activity and other cues may be affected.

Behavioural reactivity to stimuli can be affected by genetic temperament, breed, sex and age in sheep (Dodd et al. 2012), and there is considerable within-flock individual variation. As the stock habituate to the rail stimuli, their average flight distance from trains will reduce, but there may be some proportion of the flock that will always maintain a large flight distance, regardless of their experience. As herd animals, cattle to some extent, and sheep to a large extent, are allomimetic, meaning that the behaviour of one animal can cause a similar response in the rest of the group. A single nervous, highly reactive animal that startles in response to train stimuli can therefore cause disturbance in the rest of the mob. These highly reactive animals are potentially at greater risk for long-term effects on production (e.g. growth or reproduction,). Cattle can respond to stress in conspecifics and become more fearful (Boissy et al. 1998).

There is no quantified evidence to support a recommendation on a particular distance necessary to reduce behavioural impacts of rail traffic on stock. Allowing stock freedom to express behavioural responses during the adaptation phase will be important as a coping mechanism to prevent physiological impacts of stressful stimuli, during the process of habituation. While reduced behavioural responses to rail traffic over time are a good indication of habituation, it is important to consider that lack of reactivity does not mean absence of impact. An example might be a reduction in flight responses as sheep perception of the threat posed by rail traffic is down-graded, but if trains disturb animal sleep cycles, this could have a cognitive, psychological or physiological impact, without eliciting a behavioural response. Monitoring of impacts on production is recommended.

Physiological responses (acute, chronic, adaptation)

Until habituation occurs, if behavioural responses to rail traffic are insufficient to reduce the impact of the rail stimuli on sheep, then physiological responses may result, with acute, chronic and adaptive impacts. These physiological responses are not necessarily negative; for example, an acute flush of adrenaline, increasing blood pressure and heart rate, can stimulate a flight behavioural response, which reduces the risk of a more serious physical impact. However, if physiological responses to stimuli occur frequently and act for long periods, then animal health, welfare and productivity can be affected.

While the conditions of freight rail specifically have not yet been evaluated in terms of their impact on sheep and cattle physiology, it is possible to use known impacts on other species to make predictions about likely impacts of rail tracks on domestic stock.

Increased heart rate during sleep (Croy et al. 2013) responses to freight train noise and vibration have been observed in humans; however, no work has been conducted to investigate effects of rail disturbance on heart rate in sheep. Despite this lack of specific literature, much is known about the impact of various stressors on the health and physiology of sheep. An acute stressor, such as a barking dog, will elicit a plasma cortisol response; however, this response is multimodal and can vary in lactating and non-lactating sheep (Cook 1997). Similarly, repeated stress associated with handling and isolation has been shown to have longer-term impacts on immunological function in young lambs (Coppinger et al. 1991). When considering environmental stressors, the impact of disrupted rest and isolation resulting in chronic stress shows physiological impacts of HPA-axis dysregulation and judgement bias (Verbeek et al. 2019) and learning deficits (Destrez et al. 2013) in sheep.

Physiological impacts of stressors occur when animals are not able to either habituate to the stimulus (so that it is no longer stressful) or adjust their behaviour so that they can cope with a stressful stimulus. Although some acute physiological impacts are fleeting and are stimulants or associates to behavioural responses, such as, for example, increased heart rates associated with flight responses, chronic impacts of rail traffic stimuli on sheep physiology will indicate that the stock are not being offered freedom to respond to the stressful stimuli with a behavioural response. An example could be the freedom to move further away from the railway line to a location where the auditory stimulus and perceived threat of the visual stimulus is lessened. Any physiological impacts will usually become most readily evident in reduced growth or reproduction. More subtle signs of physiological responses, such as increased heart rate or impaired immune function may not be readily detected in grazed stock. Monitoring of sheep for productivity impacts, and ability of the sheep to implement behavioural adaptations so that physiological impacts do not occur, will be essential. Promoting habituation and allowing behavioural responses by initially grazing sheep where they can see and hear trains, but are not fenced close to the railway line, is recommended.

Impacts on production and reproduction

Conceptual model development

While short-term stressors and challenges are normal aspects of life, the ability of an animal to cope with these is affected by the type of stress event, the duration of the stress event, and the combined impact of stressors, known as the allostatic load (Edes et al. 2018). This is an important aspect when considering the impact of rail construction, noise, vibration and other factors.

While each of these impacts may be minimal, and largely mitigated through habituation, when combined with other stressors encountered through normal farm activities, such as mustering, weaning, shearing and other husbandry practices, it may contribute to the allostatic load experienced by the animals (Fig. 1).

Fig. 1.

Diagram adapted from Kight and Swaddle (2011), demonstrating the impacts of anthropogenic and environmental influences on allostatic load and the animal’s ability to cope.

Once the stressors have been identified, we can then try to predict the level of response and duration of the stressors on the animal (Fig. 2). For example, shearing elicits a greater stress response than intermittent rail noise and both of these stressors may occur only for a fraction of the time compared with a consistent low-level noise (Fig. 2).

Fig. 2.

A conceptual model of the relative impact of certain stressors on livestock in relation to the level and duration of stressor.

A conceptual framework for understanding how noise stimuli affects wildlife has been described in Francis and Barber (2013); however, at present similar framework is not applicable to domestic livestock. This framework describes the impact of the types of anthropogenic noise, from sudden or erratic, through to more frequent sounds with spectral overlap, within the context of the biological impacts described in the literature. The relative lack of comprehensive literature on the impact of anthropogenic noise on domestic animals limits an adaptation of this framework for this purpose at this time. We have proposed the above conceptual models outlining factors likely to affect the ability of an animal to cope with environmental stressors, including noise, in Figs 1 and 2. These figures aim to provide context for the likely/possible impact, given the limited data available. To quantify the impacts of these stressors will require empirical data, which are not currently available.

Recommendations for future research

The impact of rail corridors on livestock welfare and production presents a number of opportunities for further research because much of the existing literature is focused on wildlife. Where domestic livestock noise and vibration disturbance research exists, the stressors are not easily comparable to the stimuli that freight rail corridors generate. Further research in this area would enable the conceptual models to be refined and applied more appropriately to livestock production contexts, including to species other than the ruminants reported here. Other rail-related issues that could be potential influencers on animal production include the possible redefinition of paddock and pasture size, disruption to drainage, transfer of weeds and other biosecurity issues. These issues were not a part of the current review but should not be discounted and further investigation is required to understand the gravity of railway corridors on animal production systems and the surrounding environment.

Real-time measures, behavioural assessments, physiological measures and production impacts are recommended for long-term impact assessments of rail corridor proximity to livestock. Spatial and land-use investigations in the paddocks adjoining rail corridors would provide additional information on the role of animal choice in mitigating the impact of noise and visual disturbance on livestock behaviour. The role that habituation and generational learning plays in mitigation of disturbance over time would also warrant investigation, assisting producers and managers in their decision-making.

Assessments of mitigation strategies such as buffer zones, reduced stocking densities and vegetation would provide valuable insight into the impact of rail noise and vibration on livestock behaviour, physiology and productivity.

Additionally, investigations into the contribution of rail noise and vibration to allostatic load would provide necessary context to livestock producers, managers and researchers when managing stressors. Development of an assessment tool would be of benefit when attempting to understand the cumulative impact of stressors affecting animals during daily life and when undergoing normal farm practices and procedures.