Light therapy

To extend photoperiod in treatment groups, head-worn light masks (Equilume Ltd.) were used (Fig 1). Lights activated at dusk (studies 1, 3 and 4) or 8am (study 2) and deactivated at 11pm. In addition to natural daylength, this provided a photoperiod range from 14.5 to 17.5h, depending on the time of year.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Horse wearing a mobile light mask.

https://doi.org/10.1371/journal.pone.0227115.g001

Sample collection

Hair samples (inclusive of root follicles) were collected bi-weekly from the same area below the withers on the shoulder and stored in labelled bags. Ten guard hairs (longer and thicker than under hairs) were selected for length measurement. Hairs were fixed onto a lightly coloured surface for better visibility, straightened and measured in millimetres. Average hair length was then calculated. For weight determination, ten guard hairs were weighed, using a microbalance (Mettler Toledo UM3). Coat condition was scored visually by the same technician using a scale from 1 (summer coat: short, sleek and shiny) to 5 (winter coat: long, thick and dull) at the time of hair sampling. Rate of shedding was determined by the same technician in studies 3 and 4 by assigning scores to the amount of hair adhering to the hand after stroking three times in the direction of hair growth (1 no hair adhering to the hand; 2 a few hairs adhering to the hand; 3 plenty of hairs adhering to the hand; 4 plenty of hairs adhering to hand with additional hairs falling to the ground).

Experimental design

Study 1: This study was conducted from September 21^st to November 8^th, 2016 and used fourteen adult horses (n = 7) and ponies (n = 7) (mares and geldings) at a riding centre in County Dublin, Ireland. Animals comprised mixed breed Irish sport horses, Welsh and Shetland ponies. Both horses and ponies were used in this study to investigate potential type-related differences in treatment response. Animals were maintained in individual stables with 2-4h turnout time each day. From October 5^th onwards, animals were fitted with rugs at night to compensate for cooler temperatures. Daytime average light intensity at eye level in stables was 71.1 ± 38.6 lux (ISO-TECH ILM-01 light metre) and was provided naturally by daylight entering through windows and doorways. Overhead fluorescent lighting was seldom used during the day and never after 5pm. Animals were fed once daily with ‘Classic Fibre Mix’ (Dodson & Horrell, Kettering, UK) at 150g per 100kg bodyweight and ‘Four Seasons Horse & Pony’ (Quinns of Baltinglass Ltd., Baltinglass Ireland) coarse feed at 750g per 100kg bodyweight. Additionally, horses and ponies received 60mL and 40mL Curragh Carron Oil, plus 400g and 200g beet pulp, respectively. Hay was fed twice daily and totalled 10.5kg for horses and 7kg for ponies. Access to water was ad libitum. Animals grazed permanent pasture, which included a mix of ryegrass, clover and various weed species, for 2-4h per day. Quantities fed reflected individual requirements depending on workload, as determined by the yard’s management. The treatment group consisted of two geldings (horses) and five mares (two horses, three ponies) (n = 7) and received daily extended photoperiod via light masks from dusk until 11pm. The control group consisted of five geldings (three horses, two ponies) and two mares (ponies) (n = 7) and received no light therapy.

Study 2: This study was conducted from July 21^st to October 26^th, 2017 at the same location as study 1 and used twelve adult horses (n = 6) and ponies (n = 6). From September 19^th onwards, animals were fitted with rugs at night to compensate for cooler temperatures. The treatment group consisted of four geldings (two horses, two ponies) and two mares (one horse, one pony) (n = 6) and received daily artificial light via light masks from 8am until 11pm. The control group consisted of four geldings (three horses, one pony) and two mares (pony) (n = 6) and received no light therapy.

Study 3: This study was conducted from November 28^th, 2017 to March 4^th, 2018 and used ten Irish Sport Horses accustomed to year-round outdoor maintenance at a Sport Horse breeding farm in County Carlow, Ireland. Animals were maintained at pasture 24/7 and wore rugs for the duration of the study. They had ad libitum access to haylage and water in addition to grazing from permanent pasture, which included a mix of ryegrass, clover and various weed species. The treatment group consisted of three pregnant mares with due dates in April, May and June, one filly (female horse <4 years) and one gelding (n = 5) and received daily artificial light via light masks from dusk until 11pm. The control group consisted of three pregnant mares with due dates in April, May and June, one filly and one gelding (n = 5) and received no light therapy.

Study 4: This study was conducted from January 20^th to March 19^th, 2018 and used fourteen Connemara ponies at a stud in County Meath, Ireland. Except for four stallions, all ponies were maintained outdoors at pasture for the duration of the study. The stabled stallions received 1.5kg of grass nuts (Louis A. McAuley Ltd., Navan, Ireland) in the morning, 1kg of ‘Pegasus Cool Mix’ pellets (John Thompson & Sons Ltd., Belfast, UK) in the evening and hay twice daily, to a total of 9kg. The remaining ponies grazed permanent pasture which included a mix of ryegrass, clover and various weed species and received ad libitum hay. All animals had ad libitum access to water. No artificial light was used in stallion stables. The treatment group, consisting of two pregnant mares with due dates in March and April, one non-pregnant mare, two fillies and two stallions (n = 7), received daily artificial light via light masks from dusk until 11pm and wore rugs. The control group consisted of two pregnant mares with due dates in March, one non-pregnant mare, two fillies and two stallions (n = 7) and did not receive artificial light or rugs.

Data analysis

All data were analysed using GraphPad Prism version 6.0 for Mac. Data was first assessed for normal distribution using the D’Agostino and Pearson omnibus test for normality. Changes in the continuous variables length and weight were analysed by two-way repeated measures ANOVA followed by Sidak’s multiple comparisons test where appropriate. The scored variables of coat condition and coat shedding were analysed using non-parametric tests (Friedman test over time within groups, Mann-Whitney U test between groups at individual time points). All data are presented as means +/- SEM and P<0.05 was considered significant.

Effectiveness of extending photoperiod at the autumnal equinox

Administration of extended photoperiod and warmth, when commenced at the autumnal equinox, did not reverse the early-stage winter coat growth in stabled ponies or horses. Hairs continued to increase in length and weight, which resulted in a long, thick coat by early November. In sheep in which the rate of wool growth is seasonally regulated, changes in pelage are regulated during the transition from telogen (quiescence) to anagen (active growth) [25]. Recent interrogations of growth stages in mammalian hair follicles have identified the existence of two sub-phases of the telogen stage [11]. The authors reported that catagen (involution) is followed directly by an early, ‘refractory’ telogen phase, while at the end of telogen a ‘competent’ phase prepares follicles for transition into anagen. This supports the argument that responsiveness of hair follicles to growth stimuli changes over time. For example, A study by Plikus et al [26] found that hairs plucked from mice during early telogen took longer to re-grow than those plucked during late telogen. As hair growth increased at an equal rate over time in both groups in study 1, hair follicles had already progressed to the anagen phase prior to commencement of the study and therefore were not in the suitable state (the competent stage of telogen [26]) to be regulated by the external stimulus of extended photoperiod. This is further supported by the seasonal decline in circulating prolactin levels in horses, which begins in August and falls to only two-thirds of peak production levels by mid-Sept [5] in line with decreasing daylength. Importantly, the findings of this study highlight that waiting to initiate light therapy until mid- to late September to ensure a short, sleek coat for a late autumn or winter competition season will be ineffective in both horses and ponies.

Effectiveness of extending photoperiod one month after the summer solstice

Extending photoperiod one month after the summer solstice, before any observable changes in pelage had occurred, effectively maintained the summer coat in stabled horses, but not in stabled ponies. After 14 weeks, the hairs of treatment horses were significantly shorter and lighter compared to those of control horses. However, the hairs of all ponies grew longer, resulting in the formation of an early-stage winter coat.

As the main function of the winter coat is to insulate the body during cold climatic conditions, it is logical that temperature is an important modulator of photoperiodic influence on coat growth and was shown to be so in horses [14]. It is unlikely that growth will be delayed, or shedding accelerated, if no other mechanism, such as increased energy intake or external provision of warmth is in place to maintain core body temperature. The lower critical temperature (LCT) is a limit below which animals are required to further increase heat generation [27]. Large breeds lose less heat compared to small breeds due to a smaller heat-dissipating to heat-retaining body surface ratio [28]. This means that the LCT of horses with intact winter coats can vary between breeds and management regimes. Morgan et al. [28] estimated that ponies have an LCT of 6.9°C when fed for maintenance and 1.4°C when fed a higher energy diet for competition purposes. In warmblood horses these values are 2.9°C and -3.4°C, respectively. All animals in study 2 were fed to support low to moderate levels of exercise. Local temperatures dropped on only one occasion (mid-October) below the estimated LCT for ponies receiving a high-energy diet. However, temperatures fell below the LCT for ponies on a maintenance diet on several occasions prior to commencement of light treatment and before rugs were fitted at night. In contrast, temperatures only dropped below the maintenance LCT for larger horses [28] on two occasions (mid-October) and never below that of horses fed for competition. Additionally, a 45-minute reduction in natural daylength occurred between the summer solstice and July 21^st. The combination of this small reduction in daylength and the occurrence of temperatures below the maintenance LCT before the study commenced, may have influenced the early transition of pony hair follicles to the autumn anagen phase and the initiation of winter coat growth. In contrast, the small reduction in daylength and temperatures above the LCT for horses, may not have been sufficient to trigger the same response in horse hair follicles. They potentially remained in the late or ‘competent’ summer telogen phase, during which, according to Geyfman et al. [11] they are more responsive to stimuli such as extended photoperiod.

Seasonal adaptations, such as alterations to pelage, take time and must therefore be initiated prior to environmental conditions changing [29]. Different horse breeds are equipped with different thermoregulation strategies, including differences in behaviour, metabolism and coat [30]. As ponies are hardy, indigenous animals that are believed to have been domesticated more recently than other horse types [31], they could be more sensitive to minor changes in daylength and temperature. Another theory is that the rusticity of the pony breeds is such that they possess stronger circannual hair growth rhythms that are less amenable to phase resetting by environmental cues. The findings of this study suggest that for ponies, photoperiod may need to be maintained from the time of the summer solstice and that ambient temperatures require careful monitoring to prevent the transition to winter coat growth and to maintain the summer coat.

Effectiveness of extending photoperiod one month before the winter solstice

Administering extended photoperiod and warmth to outdoor living horses between late November and early March, resulted in shorter and lighter hair coats in treatment horses. However, differences between groups only began to emerge a few weeks after the winter solstice.

During the first weeks, hair follicles were potentially in the refractory telogen phase and did not respond significantly to the extension of photoperiod or fluctuations in ambient temperatures. Nightly temperatures frequently dropped below 2.9°C, which is the estimated LCT for horses with intact winter coats on a maintenance diet [28]. However, no changes in hair coat occurred until after January 7^th when some treatment horses started shedding. Therefore, it could be hypothesised that hair follicle cells remained unresponsive until after the winter solstice and subsequently activated follicles reacted quickly to the low environmental temperature. This may explain the sudden spike in both hair length and weight observed in both groups between January 22^nd and February 17^th. While there were no dramatic changes in temperatures recorded around this time, there was an added wind-chill caused by stronger winds that may have elicited a reactive growth response, however this is speculative at best. In treatment animals, this spike was less pronounced, suggesting that the light stimulus dampened this response.

During the last two weeks (mid-late February), daily average temperatures rose above 10°C for the first time. Daylength for control horses had increased by approximately 3.5h from that at the winter solstice. The combined effect on prolactin secretion of longer daylength and warmer temperatures [32,33] may have contributed to the observed decreases in hair length and weight in both groups at the end of the study period.

Individual variations in response to treatment were also observed. In the treatment group, two of the pregnant mares showed length reductions of between 25 and 38%, the third pregnant mare only showed a reduction of 5% whereas the gelding increased length by 20%. Similar variation was seen for hair weight. Responses to photoperiod can vary between populations and between individuals from the same population, potentially as a result of variation in melatonin responsiveness or its secretion pattern [34]. While previous studies have shown that melatonin production in horses is acutely suppressed by light [17], significant individual variation in night-time production levels were observed [35]. Additionally, different thermoregulatory requirements may cause hair growth and shedding to vary according to age, body region, breed and sex [36–38].

The efficacy of light treatment on the shedding of the winter coat may also differ significantly between outdoor and indoor housing systems due to the differing demands on thermoregulation. Heat loss per square meter of body surface area is estimated to be 20% greater in horses maintained outdoors compared to a four-walled shelter [39]. Stachurska et al. [38] found that air temperature significantly affected coat changes in outdoor living Polish Konik horses. The authors suggest that precipitation also contributes to a delay or slowing in shedding. Wind is another factor that is known to affect heat loss [28,30]. It is hard to evaluate the efficacy of the rugs used at protecting the study animals from heat loss. As piloerection (the ability of hairs to be raised, lowered or turned by hair erector muscles) can increase coat depth by 10 to 30%, it improves insulation [40]. It could be argued that rugs reduced this effect by flattening the hairs and thereby did not provide a sufficient substitute for the winter coat, potentially explaining slowness to shed despite light treatment when horses are maintained outdoors.

Similarly, nutritional factors may have impacted the effectiveness of light treatment in the outdoor living horses, as they were only provided with supplemental forage for basic maintenance. Although not perceived by staff managing the animals, this could have resulted in fat reserves being drawn upon, reducing a natural means of thermoregulation. Rugs may have only partially compensated for this. A limitation of this study may have been the failure to measure and better quantify small changes in body condition score during the study period. In conjunction with body condition scoring, future studies could look to include measurements of rump fat thickness to calculate total fat percentage, as a means of better quantifying natural insulation by subcutaneous fat.

As large numbers of breeding stock, particularly Thoroughbreds, are exposed to artificial photoperiod early in the year (to advance seasonal reproductive activity) and this has a significant effect on coat growth, a potential welfare concern worth highlighting is the turnout of broodmares on cold winter mornings while they are under a lighting regime which stimulates coat loss. Similarly, exercising horses during the cold winter months when they have a heavy winter coat may result in a significant chill factor during cool down periods following heavy sweating. Improved understanding from this and other studies investigating changes in coat growth by photoperiod manipulation at specific times of year should help better inform management regimes and reduce these health and welfare risks.

Effectiveness of administering extended light on month after the winter solstice

Administration of extended photoperiod and warmth one month after the winter solstice did not significantly decrease hair length or hair weight in indoor or outdoor living Connemara ponies. While the coat of treatment animals appeared more summer-like in condition, as determined by coat scores, and shedding of the winter coat was determined to be faster in light-treated ponies, the failure of hair parameters to differ between groups was unexpected.

It is a common observation that ponies typically grow longer, heavier and denser winter coats than horses. A likely reason for this is the fact that smaller horse breeds have a larger heat-dissipating to heat-retaining body surface ratio [28] and a heavier coat reduces heat loss in colder weather. It would therefore make sense that compared to larger breeds, winter coat growth in ponies is initiated earlier, as supported by observations in study 2. Similarly, this hypothesis may explain the reduced effectiveness of light treatment on hair changes in this study. Hair follicle transition to the active anagen phase may have been slowed by temperatures below the LCT limit, estimated to be between 1.4°C and 6.9°C [28]. As average nightly temperatures during March were still at 0.9°C, this may have dampened the effects of extended photoperiod on follicle response.

Schmidt et al. [14] found that guard hairs of Shetland pony stallions stabled between October and March at temperatures above +10°C were shorter, hairs regrew more slowly and changes in coat appeared earlier compared to those stallions maintained outdoors, despite both groups experiencing the same changes in daylength. The authors concluded that environmental temperature significantly affected pony hair coats. Similarly, Schrammel et al. [41] observed earlier shedding of the winter coat in stabled Shetland pony stallions receiving 16 h of light from December 15^th until March compared to control animals. Ten of the 14 Connemara ponies used in the current study were maintained outdoors at pasture, potentially explaining the contrasting results between studies using similarly rustic pony breeds. In the present study, shedding rate increased in both groups from February onwards but was significant only in the light-treated group.

To accelerate shedding of the winter coat in ponies effectively, it may therefore be particularly important that light treatment is combined with temperatures above 1.4°C, when feeding a high-energy diet, or above 6.9°C when feeding for maintenance. Frequent grooming may also affect the timing of shedding as it is thought to help to loosen hairs [38]. Ponies in this study were not groomed.

Advancing the transition to a short summer coat is desirable in many show breeds and is commonly achieved in competition show barns where ponies are maintained indoors, rugged, fed a high-energy diet, regularly groomed and exposed to 16 h artificial light/day. In this context, failure to achieve comparable results may be explained whereby extended daily light was administered to a cohort of an indigenous pony breed, primarily maintained outdoors and fed a maintenance diet.