|Scientific Name:||Hippocamelus bisulcus|
|Species Authority:||Molina, 1782|
|Red List Category & Criteria:||Endangered B2ab(i,ii,iii,iv,v); C2a(i) ver 3.1|
|Assessor(s):||Black-Decima, P.A., Corti, P., Díaz, N., Fernandez, R., Geist, V., Gill, R., Gizejewski, Z., Jiménez, J., Pastore, H., Saucedo, C. & Wittmer, H.|
|Reviewer(s):||Zanetti, E.S.Z., González, S. & McShea, W.J.|
|Contributor(s):||Flueck, W. & Smith, J.|
This species is confirmed to be Endangered because of an ongoing decline due to natural and anthropogenic factors (e.g. habitat loss, hunting, poaching and predation), reduction in range (area of occupancy) and because remaining populations are small and fragmented. The area of occupancy reflects the fact that the Huemul will not usually occur throughout the area of its extent of occurrence, which contains unsuitable or unoccupied habitats. Population estimates for this species are a minimum of 1,048 and a maximum of 1,500 (López et al. 1998, Povilitis 1983, Díaz and Smith-Flueck 2000). Densities of 5-8.6 Huemul/km² have been reported in well-studied groups (Díaz and Smith-Flueck 2000, Frid 1999, Grosse 1949, Wensing 2005), and based on body size, Huemul are expected to occur at densities of about five deer/km² (Damuth 1987) in their prime habitat. The population is severely fragmented (Vila et al. 2006, López et al. 1998, Díaz and Smith-Flueck 2000, Vila et al. 2006). There were 101 subpopulations recognized, 60% are found within one grid of max. 64 km² (often 10-20 individuals), 15% of groups are found within two grids of max. 128 km² (often 10-30 animals), and 8% of populations are found within areas of only 192 km² (often 10-30 animals) (Povilitis 1998, 2003).
The species is also listed under criterion C because there are fewer than 2,500 mature individuals and there is a continuing and projected decline in numbers of mature individuals (Vila et al. 2006, López et al. 1998, Povilitis 1983, Díaz and Smith-Flueck 2000).
|Previously published Red List assessments:|
|Range Description:||The Huemul is endemic to Chile and Argentina and currently only inhabits the Andes of southern Chile and Argentina (Vila et al. 2006). The subpopulation in the Nevados de Chillán area (VIII Region of Chile) has the greatest degree of isolation, with a distance of about 400 km from the nearest subpopulation to the south in Nahuel Huapi National Park (Argentina, 40º30'S, SIB-Administración de Parques Nacionales, www.sib.gov.ar). The southern part of Lanin National Park (Argentina) may still have a subpopulation, but the latest surveys did not detect the specie (Pastore et al. 2013).|
However, there were skeletal remains reported in Laguna Fria, Lanin National Park. A male was photographed in Nahuel Huapi National Park and sightings have been reported in this park (SIB-Administración de Parques Nacionales, www.sib.gov.ar). The current distribution has a high degree of fragmentation, with the coastal populations in Chilean Patagonia being the least fragmented, and the most continuous stretch of populations being in areas that are nearly void of human presence (Corti et al. 2005). The degree of isolation of subpopulations within much of the region, or its effects, remains, however, unclear.
The historical distribution range of Huemul covered a latitudinal range of 20 degrees along about 2,000 km. In southern Patagonia Huemul reached the Atlantic, and whereas it occurred in the Patagonian steppe to the north, it is presently uncertain how far east they might have reached (Díaz and Smith-Flueck 2000). Formerly the Huemul occurred along the Andes south of River Cachapoal (34°) to the Strait of Magellan (54ºS) in Chile, and from approximately 36° 50’S to southern Santa Cruz Province, Argentina (Díaz and Smith-Flueck 2000). However, even recently a Huemul was photographed at 36° 41’S near the Argentine border (Vidoz 2012). As a result of pre- and post-Columbian events, there are few historic documents of Huemul existing in extra-Andean landscapes (reviewed in Huemul Task Force 2012), but the suitability of lower elevations for Huemul is evidenced from fossil remains (Perez and Batres 2008) and anecdotal reports (see Díaz 1993, 2000).
|Range Map:||Click here to open the map viewer and explore range.|
|Population:||The species occurs in fragmented subpopulations with reported densities ranging from 0.02 - 5.66 deer/km² and averaging 1.25 (n=17) (reviewed in Díaz and Smith-Flueck 2000), however, some estimates included unoccupied areas. Wensing (2005) measured 8.64 Huemul/km² in coastal Chile. It has been estimated that no more than 1,000 animals survive in Chile (Drouilly 1983) and 500 in Argentina (Flueck and Smith-Flueck 2006), resulting from reductions of over 99% in population size and of over 50% in distribution (Redford and Eisenberg 1992). In Argentina, the remaining Huemul reside along 1,850 km of Andes in about 50 subpopulations, which are mostly fragmented (Flueck and Smith-Flueck 2006). In 2006, there was an estimated 101 subpopulations of Huemul, but 60 of these only lived within 8x8 km², whereas another 15 lived within 8x16 km². Of these 65 subpopulations, 67% are outside of protected areas (Vila et al. 2006). In Chile, most Huemul populations are concentrated in coastal Patagonian, where extensive habitat is still present in almost a continuous form. However, fragmentation increases northwards on the mainland. Only a few subpopulations of Huemul in Chile seem to be stable (Tamango and Castillo National Reserves, Torres del Paine National Park, some southern coastal subpopulations). One subpopulation in central Chile has been declining (Povilitis 1998), but trends for most subpopulations are unknown, or only became clear when local extinctions were evident. Although in some localities in Chile and Argentina sightings have increased in recent times, it has not been shown to be due to a numeric response.|
|Current Population Trend:||Decreasing|
|Habitat and Ecology:||Habitat:|
At present, the Huemul occurs mainly in the Andes mountains, from sea-level up to 3,000 m elevation, and now is found mainly at the forest edge and forests of southern beech (Nothofagus spp.). Locally, Huemul are still found in a variety of habitats: from valley flats to steep mountain slopes, from open grasslands to closed shrubby or forested habitats, inclusive of post-fire or mixed habitats. The pattern of using different habitat components has been found to vary and depends on availability, season, phenology, presences of other herbivores and predators, or disturbances, and is reflected in its broad dietary spectrum (see below). Annual home range size in some currently used habitat has been estimated to be about 350-650 ha with daily travel distances of up to 8 km, but rarely more than 5 km (reviewed in Díaz and Smith-Flueck 2000, Gill et al. 2008). Serret (2001) postulated that in Argentina the optimum habitat was the transition between the forest and the steppe, but the introduction of sheep and cattle farming and the urbanization in this ecoregion displaced the Huemul populations to the forested areas. Much earlier, Onelli observed in 1897 that most lowland areas he visited already had lost Huemul due to human presence and overhunting (cited in Gigoux 1929), and Magne de la Croix (1937) considered that human pressure was responsible for Huemul only remaining in high mountains at that time. The general displacement of Huemul from traditional winter areas, lowlands or habitats far from forested Andes has been summarized by the Huemul Task Force (2012) and Flueck and Smith-Flueck (2011a, 2012a). In the past, Huemul also occurred in completely treeless areas of the Patagonian grasslands (Díaz 1993, 2000, Frid 1994, Huemul Task Force 2012).
Huemul may form small unisex or mixed groups of 2-3 individuals in low-density populations (e.g. 1.79 ± 0.33 ind./km², Corti 2008). The main populations of Huemul are now concentrated primarily on the eastern facing slopes of the Patagonian Andes, and along Chile’s southern coast and associated archipelago. Coastal populations in the Bernardo O’Higgings National Park, Regions of Aysen and Magallanes showed a pattern for group size of seven individuals in the area of Fiordo Tempano, and groups of eight individuals at a density of 8.64 ind./km² in Fiordo Bernardo (Wensing 2005), as well as groups of 8 in Torres del Paine National Park (Guineo et al. 2008). Groups of 10 Huemul during summer were reported by Grosse (1949). In the Protected Park Shoonem of Alto Rio Senguer (Prov. Chubut, Argentina) it is common to see larger groups in autumn and winter. For instance, there are reports of groups of 11 Huemul (Díaz and Smith-Flueck 2000), and local staff report Huemul groups of up to 17. However, historically Huemul formed large wintering groups and reports of 50 or over 100 exist (e.g. Prichard 1902, reviews in Díaz and Smith-Flueck 2000 and Huemul Task Force 2012).
In Torres del Paine National Park, Magallanes - Chile, in the Grey area, Pehoé and part of Dickson Lakes there is a Huemul population of at least 60 individuals with a density of 0.7 Huemuls/km² (Guineo and Garay 2014). Within the province of Rio Negro - Argentina, population density in the lower Manso River averaged 1 ind./km² (Smith-Flueck and Flueck 1997), similar to the density reported for Lago Cochrane National Reserve (Corti 2008). The lowest density (0.4 ind./km²) was register in Nevados de Chillán, Biobío, the only population of Huemuls remaining in Central Chile (Povilitis 1978, 1998). In Fiordo Bernardo, Chile, Huemul occurred at a density of 8.64 ind./km² (Wensing 2005).
Predators of the Huemul include humans, Pumas (Puma concolor), Culpeo Foxes (Pseudalopex culpaeus) and domestic dogs. Predation can be a major limiting factor of Huemul and even threaten the viability of some remaining subpopulations when reduced to small groups (Wittmer et al. 2014). Huemul are particularly at risk from predation from Pumas (primarily known to kill adult Huemul) and Culpeo Foxes (primarily known to kill fawns) in the form of apparent competition (Corti et al. 2010, Elbroch and Wittmer 2013, Wittmer et al. 2013a). A fawn was observed being killed by a fox, then chased off by the Huemul mother (Wensing 2005, Paulo Corti unpubl.). Saucedo and Gill (2004) reported several fawns being killed by foxes. Hershkovitz (1972) described the several species of foxes, like Culpeo Fox, as dog-like or coyote-like in general appearance and habits. However, two studies of Culpeo Fox diet analysis did not find evidence for Huemul fawn in the diet even during partutition in a syntopic area (Zapata et al. 2005, Jiménez and Martinez 2010). Puma have been shown to be an important ultimate mortality factor in one subpopulation (Smith-Flueck and Flueck 2001). Huemul will escape persistent pursuers by fleeing, by seeking refuge on cliffs or by using lakes, being very efficient swimmers. When threatened by predators, Huemul may attempt avoidance by first remaining motionless or hiding. However, when discovered at close range they will rapidly flee. Huemul have the capacity to learn and have been observed to increase their flight distance in relation to disturbance activity (M.L. Thomas and O. Guineo pers. comm., Frid 2001).
Close to human settlements and near ranches, Huemul also appear at risk from predation from domestic dogs. For example, Corti et al. (2010) reported that dogs near the town of Cochrane in Chilean Patagonia were responsible for 16% of observed mortalities of marked Huemul fawns. Hunting and poaching have historically been a major limiting and exterminating factor for Huemul, the extent to which poaching continues to affect Huemul is less clear.
Studies so far showed the number of plant species in the diet as low as 11 and as high as 120 in Torres del Paine (Guineo et al. 2008), which is dependent on the vegetation community, pressure from other herbivores, predators and disturbances. On the other hand, females spent 93 % of their foraging time on only two plant species during the summer-fall season at Fiord Tempano in O’Higgins National Park (van Winden 2006). Gramineae in the summer diet was as low as 0.1% (Smith-Flueck 2003) or as high as 16% (Sierralta 2003). The results of these and other studies shows that the diet of Huemul can vary substantially from one subpopulation to another (also see Smith-Flueck 2003, Galende et al. 2005).
Huemul have high site fidelity when moving from low elevations in winter to high elevations in summer. A few historical reports refer to the vertical movements of the Huemul (Cabrera and Yepes 1960, Giai 1936, Krieg 1940, Housse 1953, Prichard 1902). In contemporary works, this seasonal movement was suggested in Chillán areas (López et al. 2004, Povilitis 1978) and confirmed in Torres del Paine (Guineo et al. 2008), Los Glaciares (Serret and Borghiani 1997) and Los Alerces National Parks (Díaz et al. 2013). Results obtained with collared animals at three study sites in Aysen, Chile, show that those Huemul undertake a modest seasonal migration or none at all. The majority of distances covered between successive locations were less than 1 km and only few over 5 km (Gill et al. 2003, 2008, Saucedo and Echeñique 2003). Previous reports of movements for unmarked Huemul ranged from 2.0 to 6.7 km for a translocated female in Torres del Paine National Park (Rau 1980) and 2.0 to 4.1 km in the Chillán area (López et al. 2000, Povilitis 1998).
Although in 1992 several factors were hypothesized to explain the lack of Huemul recovery (cattle, exotic trees, irrational forestry, exotic animals, illegal hunting, diseases, dogs, reduced numbers, loss of winter range), each when taken alone is problematic as a key explanation for the general lack of recovery (Flueck and Smith-Flueck 2006, 2011a).
|Continuing decline in area, extent and/or quality of habitat:||Yes|
|Movement patterns:||Altitudinal Migrant|
|Use and Trade:||Commercialized illegal hunting may no longer occur, or would be limited likely due to increased public awareness. Poaching is mainly by local people, occurs for various reasons: for fish bait along the Pacific coast, for food for ranch dogs and people, and simply to collect a trophy. A risk for accidental or intentional kills occurs in areas with officially permitted commercial hunting of exotic game species. Commercial sport hunting of Red Deer and feral pigs, for instance, is allowed in several areas with Huemul (including some in National Parks). In Huemul areas, these commercial activities should not be allowed; if control hunting is deemed necessary, it should be preferably a task only for the government agencies' staff -if this is not possible, it should be under strict supervision from these agencies. In any case, control should be performed only if a sound and scientifically based plan is in place. The use of dogs should be strictly regulated in all the areas of Huemul distribution. In National Parks in Argentina, a new stricter regulation is in place (Res. HD 59/13), which is being enforced successfully in some Huemul areas but not yet in others. In Chile, a new law that ban dogs from protected areas was just passed. Aside from intentional poaching, dogs accompanying local people may kill Huemul which are then utilized. Hunting with dogs for other species like feral pigs may result in by-catch kills of Huemul. Eco-tourism also exists with organized viewing Huemul in the wild (for instance http://doslagoslodge.com), but no formal management of these activities has been established.|
The widespread reduction in numbers and distributional range across the full extent of the historic range (Redford and Eisenberg 1992) was likely due to a combination from overkill, overstocking with domestic livestock (feral and controlled animals), land conversion for agricultural purposes, and building of towns and roads. This process started before the arrival of Spaniards and peaked before the first systematic accounts were made (Miller 1980), more details remain to be elucidated. Based on the chronology of human settlement, the historical impact would have been earlier in the northern than in the southern portion of the distribution, and earliest possibly in the northwestern portion, which is the central valley of Chile.
The major reduction in numbers and distribution resulted in an increased fragmentation of the population. The resulting subpopulations continue to be affected by a variety of factors, which keep their numbers low or decreasing. It is under these specific local circumstances that myriad biophysical factors can become crucial: human infrastructure, livestock grazing, farming, poaching, predators, new diseases, uncontrolled and/or feral dogs, or unregulated tourism. Huemul subpopulations from such a large variety of habitats and circumstances can be expected to be of different average physical condition, and thus might respond differently to any given factor.
One important omnipresent threat is that Chile and Argentina have weak mechanisms to protect Huemul in non-protected areas, and even inside protected areas. The many problematic activities in areas still containing Huemul are mainly related to livestock production, forest products, mining and tourism of all types including fishing and hunting.
Another omnipresent and continuous threat is lack of suitable winter range. Whereas some subpopulations have access to a sufficient altitudinal range within a small area (due to steepness of terrain), others would need to migrate over much longer distances, particularly on the eastern slopes of the Andes. In all cases, if the winter range is basically occupied by humans a subpopulation may be hindered from recovering. Most historical winter ranges on the eastern slopes are private and are used intensely for livestock production and/or hunting. This includes many protected areas including the national reserves of national parks. Notable exceptions are the Torres del Paine (Chile) and Perito Moreno (Argentina) National Parks: livestock production was removed and Huemul eventually recolonized low-lands, even into steppe areas (Guineo et al. 2008, Mirabelli 2010). However, in the former, horses and cattle still forage freely in Huemul ranges (Jimenez pers. obs.). Besides loss of winter ranges, migratory behavioral patterns have also been eliminated, as Huemul reaching those winter ranges were regularly killed.
The local circumstances of a given subpopulation would determine the need for specific research to address the problems and to find adequate counter measures (Corti et al. 2005).
Food competition from other herbivores
Currently, but certainly in the past, Huemul formed part of a natural multi-herbivore system. As in any such system, it is the combined herbivore pressure on forage that is of interest. Huemul and red deer diets in a lenga forest habitat overlapped between 42 to 62%, depending on the season (Smith-Flueck 2003). Some Huemul populations persist, even after having been in the presence of cattle production for over 110 years (Flueck 2010). The broad dietary breadth of Huemul and the variety of environments occupied historically and concurrently corroborate the co-existence of Huemul with native Pudu, Guanaco, and feral cattle during several hundred years (Flueck 2010).
In Argentina, veterinary records indicate the presence of dangerous diseases near Huemul populations, as raising livestock is a major economy for the region. In Chile, screening of 18 Huemul for infectious agents did not detect antibodies for bovine herpesvirus-1 nor Brucella abortus, only two Huemul had antibodies for bovine viral diarrhea even though this population had been exposed to sheep and cattle for at least 80 years (Corti et al. 2013). Flueck and Smith-Flueck (2012c) reported on several disease agents: adult females did not reveal ectoparasites (n=73) and testing for antibodies for foot-and-mouth disease viral antigen was negative (n=41). Fasciola hepatica was encountered in three surveys at prevalence from 9% to 50% (n=108). Taenia ovis krabbei was identified, and Cysticerus tenuicollis was found at a prevalence of 8% (n=12). Ostertagia sp., Bunostomum sp. and Dictyocaulus sp. had a prevalence of 75%, 25% and 13% (n=9), respectively. Several gastrointestinal parasites reported at low prevalence in Huemul are common in livestock and are considered commensals in domestic ruminants.
Flueck and Smith-Flueck (2008) reported on bone disease in Huemul from Argentina. Osteopathology was found in 57% of adults, 63% showed mandibular, 100% maxillary and 78% appendicular lesions. These skeletal lesions would affect the capacity for predator avoidance, suggested as an explanation for the population's low average adult age (3.1 years) and lack of recovery.
Recently, Selenium (Se) deficiency was found in the first Huemul population examined so far (Cochrane, Chile), with plasma Se deficient in 73% of individuals (Flueck et al. 2014a). The examination of skeletal remains from the same population revealed bone lesions with a pathophysiognomy similar to Argentine cases, but more severe, and importantly, occurring in a population of Se deficient Huemul (details under 'Seasonal movements' in the section Habitat and Ecology). The role Se deficiency plays in limiting Huemul is ambiguous, because there is no single definition for “healthy population” for either wild or domestic animals (Chihuailaf and Corti 2014) and Huemul has survived since the Pleistocene (Marín et al. 2013) in habitats dominated by volcanic soils lacking in Se (reviewed by Chihuailaf et al. 2014).
Corti et al. (2010) reported dogs were responsible for 16% of observed mortalities of marked Huemul, based on tracks, scats, hair, state of the carcass and predator sightings (Corti et al. 2013). Other subpopulations though have declined in the absence of dogs, and the nature of local threats has to be determined for each subpopulation. Where predation is severe, including by native predators, small Huemul subpopulations may not be able to recover without management intervention (Wittmer et al. 2014). Recovering Huemul populations affected by predation might require reducing abundant primary prey and simultaneous predator reductions, at least temporarily (Wittmer et al. 2013b). Culpeo Fox were identified as a major source of Huemul fawn morality during one radiocollar study in Southern Argentina (Corti et al. 2010), but not during a fox diet study in the same region (Zapata et al. 2005). Also, Jiménez and Martinez (2010) did not find evidence of Huemul preyed or scavenged at Tamango National Reserve when examining 409 Culpeo scats along one year in a high density breeding Huemul area. More study on the source of fawn mortality is warranted.
Corti et al. (2011) assessed genetic variability using 14 microsatellite loci in one isolated population in Chile. They sampled 56 individuals between 2005 and 2007 and found low genetic variability, whereas the low inbreeding coefficient suggested nearly random mating which would maximize the maintenance of genetic variation. As there is no apt information on Huemul, the precautionary principle indicates that increasing the size of subpopulations by all means should be a major objective.
As 63% of all subpopulations are outside of protected areas (and another 9% are partially outside), there is an urgent need to establish management plans for most of the remaining Huemul subpopulations. The Huemul is classified as Endangered in the Chilean and Argentinean Red Data Books of Vertebrates (Glade 1988, Díaz and Ojeda 2000) and is also listed in the Appendix I of CITES and UNEP/CMS Conventions. This species has been protected by law since 1929 in Chile and 1989 in parts of Argentina (Díaz and Smith-Flueck 2000). In order to promote Huemul conservation, six binational workshops have been conducted since 1992. In 2001, both countries developed their national Huemul conservation plans (Chile began with an update in 2007), but funding constraints have thus far impeded their implementation. The species occurs in 13 Chilean national parks and reserves (Corti et al. 2005), and details about populations occurring outside of protected areas are either barely known or unknown. In Argentina, Huemul occur in five national parks plus several provincial reserves. Both countries also have some private reserves with Huemul populations. Nevertheless, only 28% of the 101 identified subpopulations in both countries were found within existing protected areas, 63% were located outside of protected areas, while 9% were found partially within protected areas (Vila et al. 2006).The public system of protected areas contains extensive habitat formerly and currently still used by Huemul, making it a prime candidate for contributing significantly to Huemul recovery (Flueck and Smith-Flueck 2006). An urgent task is thus to find ways to extend the implementation and control of existing policies established for protected areas (Martin and Chehebar 2001, Rusch 2002).
Some conservation measures currently prioritized by government agencies of both countries include: increase efforts to obtain more information on Huemul subpopulations such as the current distribution, abundance, and threats; encourage more effective protection of the identified subpopulations (emphasizing an increase of protection in the VIII and XI regions of Chile and in the Province of Chubut of Argentina); promote the creation of private protected areas with presence of Huemul (or suitable habitat) to facilitate connectivity and dispersal; promote training to improve local skills in wildlife management and monitoring techniques; and encourage educational activities and media campaigns to raise awareness about the Huemul´s status (Smith-Flueck et al. 2014). Achieving these measures would be more likely through coordination (also binationally) among the several administrative jurisdictions, private owners, and local communities.
The present state of Huemul with its high degree of spatial fragmentation, small subpopulation sizes, and low total number (Vila et al. 2006) indicates that the situation is not viable, and may be less so in Argentina with only a few hundred remaining Huemul. The trend follows a continued loss of subpopulations, i.e., a reduction in numbers and area of occupancy, even within national parks (Franke 1949, Povilitis 1998, Serret 2001). The goal is the species’ recovery (e.g. Argentina National Huemul Conservation and Recovery Plan, 2001), which implies an increase in numbers and distribution to a viable level. For recovery to succeed, threats to individual remaining Huemul subpopulations need to be understood (Corti et al. 2005). Making decisions about corrective measures should be based on understanding these threats, and would take conservation beyond the creation of protected areas. Basic information on the species and on most subpopulations is very scarce, which complicates interpretations of some observations. A recovery strategy for Huemul must also include ex situ tools, as for many of the questions, the most efficient, and sometimes only, approach would be controlled studies of semi-captive animals (a successful captive center occurs in Chile since 2005, Vidal et al. 2011). It would also provide opportunities to study re-introduced groups (stemming from the captive breeding center), taking an adaptive management approach (Flueck and Smith-Flueck 2006, Armstrong and Seddon 2007). The Huemul captive centre in Chile is planning their first re-introduction of five Huemul for later in 2015. In consequence, several additional conservation measures surfaced that need to be incorporated: 1) Scientific research should be facilitated by all agencies involved and for any existing Huemul subpopulation; 2) Centres with semi-captive Huemul for studies and re-introductions function as a valuable recovery tool and should be facilitated, particularly when non-competing funds are available. Efforts should be directed to gain a maximum of information through research under controlled conditions and through reintroductions; 3) Once subpopulations are known to be recovering numerically and spatially, they should become a research focus in order to determine the reasons for the recovery; and 4) The groups assigned to coordinate a centralized and country-wide conservation effort based on Huemul National Plans (“Comisión Nacional” in Chile, “Comité de Gestión” in Argentina) need to adopt a policy of transparency.
|Citation:||Black-Decima, P.A., Corti, P., Díaz, N., Fernandez, R., Geist, V., Gill, R., Gizejewski, Z., Jiménez, J., Pastore, H., Saucedo, C. & Wittmer, H. 2016. Hippocamelus bisulcus. The IUCN Red List of Threatened Species 2016: e.T10054A22158895.Downloaded on 23 March 2017.|