SciELO - Scientific Electronic Library Online

vol.13 número2First report of mites and fleas associated with sigmodontine rodents from Corrientes Province, ArgentinaMicromamíferos terrestres de Puerto Lobos, Chubut, Argentina índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




  • No hay articulos citadosCitado por SciELO

Links relacionados


Mastozoología neotropical

versión impresa ISSN 0327-9383versión On-line ISSN 1666-0536

Mastozool. neotrop. v.13 n.2 Mendoza dic. 2006


Unusual mass mortality of feral horses during a violent rainstorm in Parque Provincial Tornquist, Argentina

Alberto L. Scorolli, Andrea C. Lopez Cazorla, and Lidia A. Tejera

Cátedra de Zoología de Vertebrados, Departamento de Biología, Bioquímica y Farmacia, Universidad Nacional del Sur, San Juan 670, 8000 Bahía Blanca, Argentina. <>

Key words. Die-off. Equidae. Equus. Population management.

   Extreme climatic events may seriously affect animal populations as well as their viability (Shaffer, 1987; Boyce, 1992). Several cases of mass mortality in wild equids caused by climatic factors have been reported; e.g. the death of plain zebras (Equus burchelli) (Walker et al., 1987) and of wild asses (E. asinus) (Saltz, 2002) which were both provoked by drought, and the death of mountain zebras (E. zebra) caused by snow storms (Penzhorn, 1984). The few reports available to date on the die-off of wild horses in other countries attribute this phenomenon to prolonged droughts (Berger, 1983; Dobbie et al., 1993), exceptional snowfall (Berger, 1983; Garrott and Taylor, 1990), and severe winters (Welsh, 1975; Garrott and Taylor, 1990). The purpose of the present research is to describe the mass mortality of feral horses (E. caballus) which occurred in Tornquist, Argentina, on 9-10 November 2002, caused by a violent rain storm accompanied by strong winds and lightning.
   The Parque Provincial Ernesto Tornquist is located in the south of Buenos Aires Province, Argentina (38 º 00'- 38º 07'S and 61º 52'- 62º 03'W). This reserve covers 67 km² of hilly grasslands, ranging between 450 and 1175 m. The climate is temperate and humid with a mean annual rainfall of 800 mm. Rains fall mainly in spring with a second peak in autumn, and snowfalls are occasional and, in general, light. The dominant typical vegetation is grassland steppe dominated by Stipa and Piptochaetium spp. (Cabrera, 1976).
   A group of 10 horses of Criollo breed was introduced in 1942 and became feral. At present the herd occupies a fenced-off area of approximately 20 km² (Scorolli, 1999). The highest number of feral horses recorded in 2002 was 650, equivalent to a density of 32.5 horses km², one of the largest recorded for feral horses to date. Although there is no official management strategy for the herd, there is a general consensus aiming at controlling its population size without eradicating the herd.
   The data analysed for this study were collected on 12-13 November 2002, i.e. two days after the reported storm. To accurately register the number of dead animals, an intensive search was made on foot to cover the entire study area, carried out during two successive days, totalling 20 hours. Every dead animal was recorded, when possible noting sex, age class (foals, yearlings, 2-year-olds, and adults), coat, and face and leg markings. The location of each dead animal was marked on a simplified topographical map. The posture and other external symptoms were also recorded. Mortality rate was estimated by each age and sex class as a percentage of the estimated total for the same age-sex group in the population. It was not possible to estimate mortality rate among males older than one year; there was no precise estimation since at this age animals generally disperse from their natal herd.
   Body condition was estimated by direct observation using Rudman and Keiper's rump index (1991) and was compared with the body condition of the population a week before the storm by means of the non-parametric Mann-Whitney-U test (Zar, 1984). Potential changes in social structure of the feral horse population were based on individual identification data collected during a long-term study initiated in 1995.
   A total of 235 mm of rain fell during the 48 hours that lasted the storm, exceeding all the records of the century [data provided by Satelmet (regional meteorological service) and Centro de Estudios de Recursos de la Zona Semiárida (CERZOS)]. There were numerous lightning discharges and a southerly wind blew at speeds of 50 to 80 km/h for more than 36 hours. The day temperature decreased abruptly from a maximum of 26 ºC and a minimum of 11 ºC registered on the day prior to the storm, to 11 ºC/6.5 ºC and 8 ºC/6 ºC on 9-10 November, respectively.
   A total of 193 horses were recorded dead within a narrow strip 200 m wide and 4 km long, parallel to the perimetral fence of the reserve and perpendicular to the prevailing wind direction during the storm. This area, at an altitude of approximately 550 m, is the lowest occupied by the horses, covering approximately 30% of the reserve. It has no sheltered places and readily collects rainfall. The dead animals were found in groups, in some cases piled together. Numerous signs of prolonged trampling were observed and some horses were partially buried in the mud. Despite a thorough search no dead horses were found along the hillsides and slopes which make up 70% of the reserve.
   None of the surviving horses observed during the search, including foals as young as 15 days old, showed visible signs of physical damage, suggesting that the majority had found adequate shelter during the storm.
   Of a total population of 700 horses, 193 were killed during the storm, and of these 12 were foals, 49 yearlings, 46 2-year-olds, 74 adults, and 11 not identified. The mortality rate of females was 0.41 in foals, 0.48 in yearlings, 0.58 in 2-year-olds and 0.31 in adults. In males the mortality rate was 0.33 in foals, and 0.57 in yearlings. Mortality was clearly female-biased only in adults (62 dead adult horses were females, only 12 were males). Of these, two were stallions who owned a band and were older than 12 years of age. The social structure changed little, all 68 known bands lost individuals to a greater or lesser degree, but none was totally eradicated.
   Of the total number of dead animals, 133 were found in a recumbent posture, three of them with their necks curved in an 'S' shape. The remaining 60 were found kneeling, 58 with similar neck curvature. A considerable number of horses presented symptoms of acute stress, 37 had blood in their nostrils, 8 had mucous, and 5 had froth, while 15 presented their tongue hanging partially out of their mouth.
   Body condition does not seem to have had an influence on mortality. The observed average body condition by age and sex class was, in general, the same as the average body condition recorded in the population the previous week. Only mean adult male body condition (1.94±0.46, N=12) was significantly lower than the population as a whole (3.38±0.50, N=75) (U= 8.5, p<0.01). Conversely, dead adult females had significantly better mean body condition (2.07±0.64, N=59) than the population (1.83±0.76, N=142) (U=3307, p<0.05).
   The total number of dead animals recorded in the Tornquist die-off was higher than that reported in Great Basin (Berger, 1983), and similar in magnitude to those reported in Sable Island (Welsh, 1975), in Pryor Mountain (Garrott and Taylor, 1990), and from U.S. Bureau of Land Management files (Berger, 1983). This could partly be due to the high density of the Tornquist feral horse population at the time of the storm. Penzhorn (1984) claims that during a storm mountain zebras instinctively turn their rump into the wind, thus protecting their heads. This may explain why some of the dead horses were clustered in the lower area, near the northern border of the reserve, as the wind was blowing in that direction. The 11-thread wire fence around the perimeter of the reserve may have proved an insurmountable barrier for the gathering horses.
   The great amount of runoff flowing towards that area may have either dragged them there or made it difficult for them to escape.
   The fact that only two band stallions died may suggest that the rest of the stallions and their harem-bands or parts of them, moved to a safer place before the die-off. Mortality could not be clinically confirmed as no necropsies were performed. The most plausible causes of mortality seem to be either electrocution or fulguration produced by lightning, which could have been transmitted through water in the low-lying, flooded area where the horses had gathered. Veterinarians consulted believe that the observed symptoms corresponded to sudden death, probably caused by lightning strike (Members of the Asociación Argentina de Veterinaria Equina, personal comments). Another hypothetical factor is a combination of both hypothermia and acute stress due to the joint effect of rain, wind, and cold during a prolonged exposure to the storm. No similar die-off events have been reported after hurricanes for populations of white-tailed deer (Odocoileus virginianus) (Labisky et al., 1999) or European roe deer (Capreolus capreolus) (Gaillard et al., 2003). The present study reports the first evidence of mass mortality in feral horses, and large mammals in general, caused by a rainstorm with strong winds and lightning. After the storm, the feral horse population was substantially reduced in size and is therefore expected to increase in size at a higher rate than previous years. On the other hand, the change in the proportion of sexes in favour of males may temper any increase in population growth rate. It can therefore be concluded that under climatic conditions and in populations similar to those of Tornquist, this type of mass mortality, although unusual, should be considered as part of environmental stochasticity when making management decisions.


   The authors are grateful to Diego Birochio for reporting the die-off; to local guide Rubén Gonzalez for his assistance in the field; to vets Randall Arms, Arturo Baldini, Oscar Calvete, Hugo Funtanillas, Carlos Taffarel, and Jaime Tasso for their contributions as members of the forum of the Asociación Argentina de Veterinaria Equina in relation to the possible cause of mortality; to Daniel Dodero (Satelmet) and to Prof. Diego Bentivegna (CERZOS-UNS) for providing meteorological data and to Dr. Ricardo Ojeda for helping with bibliography. This work was supported by a grant (PGI 24/B074) from the Secretaria de Ciencia y Tecnología of the Universidad Nacional del Sur, Argentina.


BERGER J. 1983. Ecology and catastrophic mortality in wild horses: implications for interpreting fossil assemblages. Science 220:1403-1404.         [ Links ]

BOYCE MS. 1992. Population viability analysis. Annual Review of Ecology and Systematics 23:481-506.         [ Links ]

CABRERA AL. 1976. Regiones Fitogeográficas Argentinas. Pp. 1-85, in: Enciclopedia Argentina de Agricultura y Jardinería. Tomo 2, Fascículo 1. (LR Parodi, ed.). ACME, Buenos Aires.         [ Links ]

DOBBIE WR, DMcK BERMAN, and ML BRAYSHER. 1993. Managing vertebrate pests: feral horses. Australian Government Publishing Service, Canberra.         [ Links ]

GAILLARD JM., P DUNCAN, D DELORME, G VAN LAERE, N PETTORELLI, D MAILLARD, and G RENAUD. 2003. Effect of Hurricane Lothar on the population dynamics of European roe deer. Journal of Wildlife Management 67:767-773.         [ Links ]

GARROTT R and L TAYLOR. 1990. Dynamics of a feral horse population in Montana. Journal of Wildlife Management 54:603-612.         [ Links ]

LABISKY RF, KE MILLER, and CS HARTLESS. 1999. Effect of Hurricane Andrew on survival and movements of White Tailed Deer in the Everglades. Journal of Wildlife Management 63:872-879.         [ Links ]

PENZHORN BL. 1984. Observations on mortality of free-ranging Cape mountain zebras Equus zebra zebra. South African Journal of Wildlife Research 14:89-90.         [ Links ]

RUDMAN R and RR KEIPER.1991. The body condition of feral ponies on Assateague Island. Equine Veterinary Journal 23:453-456.         [ Links ]

SALTZ D. 2002. The dynamics of equid populations. Pp. 118-123, in: Equids: Zebras, Asses and Horses. Status Survey and Conservation Action Plan (PD Moehlman, ed.). IUCN/SCC Equid Specialist Group. IUCN (The World Conservation Union), Gland, Switzerland and Cambridge.         [ Links ]

SCOROLLI AL. 1999. Demografía y áreas de actividad de caballos cimarrones en el Parque Provincial Ernesto Tornquist. Magister thesis. Universidad Nacional del Sur.         [ Links ]

SHAFFER M. 1987. Minimum viable populations: coping with uncertainty. Pp. 69-86, in: Viable populations for conservation (M Soulé, ed.). Cambridge University Press, Cambridge.         [ Links ]

WALKER BH, RN EMSLIE, N OWEN-SMITH, and RJ SCHOLES. 1987. To cull or not to cull: lessons from a southern African drought. Journal of Applied Ecology 24:381-401.         [ Links ]

WELSH D. 1975. Population, behavioral and grazing ecology of the horses of Sable Island, Nova Scotia. Ph.D. thesis. Dalhousie University.         [ Links ]

ZAR JH 1984. Biostatistical analysis. Prentice-Hall, New Jersey.         [ Links ]

Recibido 7 julio 2005.
Aceptación final 17 febrero 2006.

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons