The project has been concluded. First data of the project was published in the Journal of Equine Veterinary Science. See under publications “Brand Anna, Arno Lindner, Hartmut Gerhards, Mathilde Valenchon, Odile Petit (2021) Measurement of Endurance of Untrained Five-Year-Old and Six-Year- Old Horses Raised in Multi-age Herds on Pasture. Journal of Equine Veterinary Science;”.

Anna Brand, Arno Lindner, Odile Petit, Mathilde Valenchon


Since 2016 and until 2018/2019


This study is part of a large project aiming to understand the influence of an enriched natural environment on the development of young sport horses. The aim is to assess the social and behavioural development of those young horses and their level of welfare. In addition, the development of the endurance and gait characteristics of the horses is followed. To do so, we are following more than 30 horses from the age of 4 to the age of 6 during three consecutive years (10 horses of each age group; start of the study was 2016). The study is conducted in a breeding farm (Les Dannes, France) that developed management methods that are closer to natural living conditions than traditional practices in the domain of equine sports. At the end of the study, we will search for relationships between behavioural parameters and the horses’ physical performances.
One hypothesis of the study is that the behavioural results of these horses will show them to be more relaxed and manageable than horses of horses managed conventionally. The other hypothesis is that over time the endurance and gait of the horses will change (improve?) because of growth and the free movement at all times.

Material and Methods
  • The farm breeds and raises competition horses (French Saddlebred). Horses are raised in conditions close to naturalistic conditions: large social groups, outdoor life (pastures, large stabling), and low-energetic food only (i.e. grass, fodder, no concentrate pellets). The farm hosts near to 400 horses (foals, yearlings, young horses, broodmares, stallions, and board horses etc.). Most of the horses are always in large multi-age social groups (20-60 individuals).
  • Horses aging from 4 to 6 years old are being studied from 2016 on and until 2018. All horses were born at the farm and raised in large multi-age groups.

Study design

  1. Social behaviour
    The behavioural observations consist in observing the horses daily in their social group during two consecutive months in summer. During each observation session, the social interactions (agonistic and affiliative) are recorded continuously. In addition, the activity (e.g. walking, grazing, resting) and the inter-individual distances between horses are recorded every 15 min (instantaneous scan sampling, Altmann, 1974). Using this protocol, each horse is observed during a total of at least 90 hours and at least 300 scan samplings per horse are recorded. The social parameters will be used to assess the development of the social network (affiliative network and hierarchy) and social competences of each horse according to its age. They will also be used to measure the level of aggressiveness. The other behavioural parameters will enable us to estimate the activity budget (i.e. the time spent in each activity) of each horse. Then, we will be able to test whether these activity budgets meet the welfare criteria. To do so, these activity budgets will be compared with activity budgets from data on semi-free ranging and feral horses (based on literature). We will also verify the presence (or not) of stereotypic and abnormal behaviours. Eventually, we will test the correlations between the activity budgets (e.g. time spent walking) and the physical performance of the horses.
    This observation protocol will be repeated each summer during 3 consecutive years. Therefore, at the end of the 3rd year of the project, we will have longitudinal data to study the behavioural development of one cohort from their 4th to their 6th year of life. In addition, the first year of the project (2016), we have been able to compare 4 years old vs. 5 years old vs. 6 years old horses.
  2. Personality tests
    We assess the horses’ personality using standardized behavioural tests that have been developed by Léa Lansade and her colleagues in Tours, France (Lansade et al. 2016). We focus on the dimension of fearfulness. Those tests are conducted once at the age of 5 and 6, when the horses are familiarised with basic human manipulations. The tests consist in submitting the horses to three fear tests: approaching a novel object, crossing a plastic tarpaulin to access food, and reacting to the sudden opening of an umbrella. The behaviour of tested animals is continuously filmed and recorded by an observer. The horses are tested in the indoor arena of the farm under standardized conditions.
  3. Endurance
    Horses are submitted to a standardized exercise test (SET) to determine their v4 and v180 (v4 = speed at which a blood LA of 4 mmol/l is determined; v180 = speed at which a heart rate of 180 beats/minute is determined; Persson 1983; Lindner 2010). The SET is performed by the horses ridden trotting and cantering in circles in a 30 x 50 m indoors sandy riding arena. The increase in speed from interval to interval was designed to continuously raise blood LA from the level before exercise and after warm-up to ≥ 4 mmol/l LA in not less than 4 intervals and maximum 5 intervals. The duration of interval 1 to 4 is just above 5 minutes, i.e. when the horses finish the next lap after 5 minutes trotting or cantering at the respective speed. The 5th interval lasts close to 3 minutes only. This provided at least 4 values to describe the blood lactate running speed relationship (BLRS). Intervals are separated by a period of about 1 min with the horses walking. Time is monitored by an observer with a stopwatch. Within 20 s after each interval, the jugular vein is punctured with a 23 g needle and blood collected into 1 ml heparin coated syringes. Blood LA is measured immediately with the hand-held LA measuring device Accutrend Plus System, Roche, Mannheim, Germany. With the blood LA and running speed for each interval of the SET, v4 is calculated by exponential regression analysis (Galloux 1991). Horses are fitted during the SET with a commercial HR monitor (Polar M400, Polar Electro OY, Kempele, Finland). The mean HR (HRmean) of each interval is calculated with the values representing a continuous plateau within the values of an interval. The HRmean and speeds of every interval allow for the calculation of v200 by linear regression analysis.
  4. Gait analysis
    The gait analysis is done with a commercial accelerometric equipment (EquimetrixTM, Centaure Metrix, Fontainebleau, France). The gait variables measured are stride frequency, stride regularity, dorsoventral displacement, dorsoventral power, lateromedial power and propulsion power at walk, trot and canter.