Locomotor activity analysis based on habitat, season and time of the day in Polish Konik horses from reserve breeding using the Global Positioning System (GPS)

The history of the Polish Konik breed is linked to that of the East European Tarpan horses (Equus caballus gmelini Ant., forma silvatica Vet.), which became extinct in the 18^th century. However, some evidence exists that they were crossbred with local primitive domestic horses. Reconstruction of the Tarpan-like horse was initiated at the beginning of the 20^th century (Pluta et al., 2016). For this purpose, a group of Koniks most resembling Tarpans was purchased from local farmers and located in the Bialowieza Forest Reserve.
       
The Polish Konik, as the only native horse breed originating from the Tarpan, is predisposed to live under natural environmental conditions, in the so-called reserve breeding. This property stems from Professor Vetulani’s theory, according to which a forest variety separated from Tarpans living on the East European steppes. It was named Equus cab. Ant., forma silvatica Vet. (Vetulani, 1933a, b). Under such breeding conditions, natural behavioral patterns inherited from ancestors emerge in animals voluntarily or instinctively (Kaleta, 2007; Mills and Nankervis, 1999). A drive for movement is one of the most important needs of horses, which enabled the formation of this species during evolution. Under natural conditions, horse movement results mainly from the need to satisfy hunger and thirst, migration or escape from danger as well as social relations and pack hierarchy (Rehm, 1981). Horses kept in stables are deprived of such possibilities. Lack of physical activity not only results in a series of problems associated with the locomotor system (such as tendon weakness, reduced bone density, abnormalities in the hoof structure and mechanics) (Nielsen et al., 2000; Porr et al., 1997), but also negatively affects the circulatory and respiratory systems (Vervuert and Coenen, 2002).
       
Studies on the locomotion of horses under field conditions are quite challenging. One method of its monitoring is to use the Global Positioning System (GPS) (Ramesh and Rana, 2016; Sahu et al., 2015; Thilagam and Sivasamy, 2013), which has already been applied to horse movement analysis by several authors (Bat-Oyun et al., 2018; Hampson et al., 2011, 2013; Hebenbrock et al., 2005; Hennig et al., 2018; Sato et al., 2017). Therefore, the aim of the present study was to analyze and understand the locomotor activity of horses living under near-natural conditions using the GPS.
       
The study was conducted in the Zagroda breeding reserve in 2018. The reserve is located in the Kliniska Forest District near Szczecin. It encompasses an area of 700 ha comprising mainly forests and meadows. The observations were carried out on the pack led by the stallion Nagaj (consisting of 15 Polish Koniks) using the GPS transmitter (Ecotone, Gdynia, Poland). The device was attached to the neck tethering of the broodmare (the so-called Wolf collar) and sent data on pack location every hour. During the same period, two more packs remained in the Zagroda nature reserve: one belonging to the stallion Nokturn (ten horses) and the second one belonging to the stallion Litwor (eight horses).
 
Three habitats were distinguished in the present study:
1. Forest, including the area of old-growth forest, coppice, forest plantations, fire-lanes and clearings.
2. Meadows, including the area of meadows (approximately 70 ha) located along the Ina River.
3. Hay, including the area at the boundary between meadows and forest, where hay was provided to Koniks.

To investigate spatial and temporal trends in the migration of Polish Konik horses within the habitats, Ecotone software was used. Based on these data, the distance travelled over one hour was determined. It was subsequently expressed in meters, in the so-called straight line.
       
The distance was recorded for 24 consecutive hours (over the period of 30 days during each season), which were subsequently grouped into the following times of the day:
 
1. Night-morning, from 00.00 to 06.00 hours
2. Morning-midday, from 06.00 to 12.00 hours
3. Midday-evening, from 12.00 to 18.00hours
4. Evening-night, from 18.00 to 24.00 hours
 
       
In addition, season was taken into account in the present study. The measurements were collected in winter (January) and spring (April and May). The data were analyzed using the following model:

 
 

where:
m is the overall mean, a_i is the effect of season (winter, spring), b_j is the effect of habitat (forest, meadow and hay), c_k is the effect of time of the day, (ab)_ij is the effect of the interaction between season and habitat, (ac)_ik is the effect of the interaction between season and time of the day, (bc)_jk is the effect of the interaction between habitat and time of the day, α is the regression coefficient for the day temperature, tM_ijkl is the day temperature, β is the regression coefficient for the night temperature, tE_ijkl is the night temperature, e_ijkl is the random error.
       
All the computations were performed using Statistica 13.1 software (Dell Inc., Tulsa, OK, USA).
       
The study revealed a statistically significant effect of habitat (p<0.0000) and time of the day (p<0.0000) on the locomotor activity of Polish Konik horses and the distance they had traveled. It can be stated that one of the factors affecting an increased locomotor activity of Polish Konik horses was the availability of the feed base. The Koniks traveled the longest distance in search of feed in the forest, i.e. 272 m per hour on average, which is 6528 m per day. This distance differed significantly from that covered in the meadow (107 m per hour on average; 2568 m per day) and hay (62 m per hour on average; 1489 m per day) habitats (Table 1). In the work by Golonka (2009) on Polish Konik horses from the Popielno Nature Reserve (using the amount of time utilized by horses for each activity and not the distance covered by them), adult horses spent an average of 54.9 min per day (3.8% of the whole day) moving. The study on Przewalski’s horses in Mongolia (Kaczensky et al., 2008) using the ARGOS and GPS systems, showed that the average daily straight-line distance between the consecutive days of observation was 3.5 km, whereas the mean daily distance traveled by feral horses in the USA (recorded using the same method) was 9.0 km (Hennig et al., 2018). In the paper by Hampson et al., (2010b) on the movement patterns of domestic Quarterhorse×Australian Stockhorses, Quarter horses and feral horses, the average distances covered by domestic horses were generally greater in larger paddocks (4.7 km per day for 0.4 ha, 6.1 km per day for 4.0 ha, and 7.2 km per day for 16.0 ha). Feral horses, living in a 4,000-ha paddock, covered the greatest daily distance (17.9 km), whereas horses kept in the 6×6 m yard moved only an average of 1.1 km daily. Similar behavioral patterns have also been observed in other species (Taskin and Karadavut, 2017; Thakur et al., 2016). In the next study by Hampson et al.,  (2010a) on feral horses in “Outback” Australia, the average daily distance traveled by all animals included in the study was 15.9 km. The central Queensland horses moved an average of 16.8 km daily and the central Australian horses 14.7 km, but the difference was not statistically significant. The differences in daily distance depending on the physiological state of mares kept for milk production in Mongolia were analyzed by Bat-Oyun et al., (2018) using the GPS method. The daily cumulative distance (approximately 0.6 – 1.0 km per hour) and the daily maximum linear distance (about 3.1 – 3.8 km) differed significantly between the milking and non-milking periods. Finally, Claudi and Hoy (2013) studying the horses of different breeds using the GPS found large individual differences in the daily distance (5.1 - 10.3 km). It was not influenced by age or sex, but considerable inter-breed differences existed (Friesian horses - 9.5 km, Warmblood horses - 8.1 km and Connemara-ponies - 7.7 km each day).
 

       
No statistically significant differences in the distance traveled by Koniks were found between seasons, i.e. winter (166 m on average) and spring (192 m on average) (Table 1). However, a significant interaction between season and habitat was observed in our study. It most probably resulted from the fact that Koniks spent more time staying near the hay feeder (feeding) in winter and were not forced to move in search of feed, as during the warmer period.
       
Clear and statistically significant differences in the Koniks’ locomotor activity between times of the day were noticed in the present work (Table 1). Horses covered the longest distance (248 m per hour, which is 1488 m during the six-hour period) in the midday and evening hours. It differed significantly from those traveled during other times of the day (Table 1). Similar distances were recorded in the evening and night (134 m per hour) as well as night and morning (124 m per hour) hours (804 and 744 m during the six-hour period, respectively). Horse activity increased to 209 m per hour (1254 m during the six-hour period) with the onset of the day (morning and midday hours) and, consequently, during the midday and evening hours (247 m per hour). In the study by Jodkowska et al., (2015), Polish Konik horses migrated most intensively between 06.00 and 20.00 hours with a slightly lower activity from 14.00 to 18.00 hours both in April 2011 and 2012. However, the overall frequency of location changes (meadows, forests) was higher in 2011. Moreover, the migration range (classified into three categories: small, medium and high) was larger in 2011, when the medium and high classes predominated, whereas in 2012 the medium category was the most frequent one.
       
No statistically significant effect of temperature (either day or night) on the distance was found. Also, no significant influence of the interaction between time of the day and season was observed, although an average locomotor activity was somewhat higher in spring (Fig 1A). Statistically significant interactions between time of the day and habitat were noticed. For each time of the day, the distances covered in the meadow and hay habitats were different from those traveled by horses in the forest (Fig 1B). The interaction between season and habitat showed that statistically significantly shorter distances were covered in the hay habitat in winter (78 m per hour) than in spring (125 m per hour) (Fig 1C). The distances in the remaining habitats did not differ significantly.
       
 

 
The course of locomotor activity during the day showed a marked increase in the morning and midday as well as midday and evening hours in winter and quite similar trends in spring, although in the latter case the activity started earlier and ended later (Fig 2). Polish Koniks, as free-living horses, kept in a nature reserve stay in the area occupied by the pack (the pack territory). The territory should be large enough to meet their nutritional requirements. The Koniks’ locomotor activity mainly results from the accessibility of resources needed for life, i.e. associated with the so-called Five Freedoms of animal welfare (Jaworski, 2003). The most important factor affecting daily distance is the availability of grass and water. An average covered distance is about 6 km; however, under ideal nutritional and living conditions, it can amount to only 2 – 3 km, with the range from 2 to 17 km (Frentzen, 1994; Hoffmann et al., 2011; Rehm, 1981). Similar values were obtained in the present study, in which Polish Konik horses covered a daily distance of more than 6 km when searching for feed in the forest habitat, but only about 2.5 km and 1.5 km when staying in the meadow and hay habitats, respectively. According to Vervuert and Coenen (2002), horses spend approximately two-thirds of their locomotor time on feed-searching. Smugała and Pikula (2014) showed that Koniks with more difficult access to feed base increased their distance in search of feed in winter, but they clearly preferred meadow ecosystems in spring, which was associated with the appearance of the young sweet shoots of the common reed. Similar results were obtained by Smugała and Pikuła (2013) in their study on the use of telemetry for the analysis of the Polish Konik pack migration, in which horse movement was also mainly determined by the availability of the feed base. Meadows were not attractive to horses in winter due to ice cover, although they spent whole days grazing in them during spring.
 

       
Brinkmann et al., (2012), investigating locomotor activity and physiological indices in ten Shetland pony mares, also found a statistically significant effect of season on the locomotor activity of horses, which plays a special role in maintaining good health and efficiency of physiological functions. According to Frentzen (1994), the so-called functional areas should be created in stable-pasture systems in order to force horses to move. Frentzen (1994) noticed that horses housed in stables equipped only with outside pens traveled about 1.8 km, whereas animals staying in those with additional functional areas covered a distance of up to 5 km. It seems that feeding Polish Konik horses hay during mild winter and spring conditions in the studied Zagroda reserve was unnecessary, since it resulted in their decreased activity in the hay habitat. A large effect of pack size on locomotor activity was observed by Hoffmann et al., (2011), who investigated different horse packs using pedometers and found a significant difference in the number of steps taken by horses from a small pack (eight horses, 82 steps per 15 min) and a three-times larger pack (23 horses, 149 steps per 15 min). The cited authors explained this difference by greater interaction among horses, social relations and the establishment of pack hierarchy.

The present study showed that the basic factor affecting the locomotor activity of Polish Konik horses kept under the reserve system was the availability of the feed base. Similar weather conditions in winter (very mild) and spring did not result in an increased locomotor activity. The clear differences in this activity were observed between the times of the day, with its highest level in the morning and evening.