Biodiversity loss is one of the most severe environmental problems, with thousands of species having gone extinct since the start of the 20th century and hundreds of thousands of populations being lost as a result of human activities (Ceballos et al., Reference Ceballos, Ehrlich and Dirzo2017a). The rate of extinction of vertebrate species in this period has been up to 100 times faster than in the last 2 million years (Ceballos et al., Reference Ceballos, Ehrlich, Barnosky, García, Pringle and Palmer2015). Extinctions of birds and mammals have been documented more thoroughly than extinctions of fishes and invertebrates, which have often been overlooked (e.g. Dirzo et al., Reference Dirzo, Young, Galetti, Ceballos, Isaac and Collen2014; Régnier et al., Reference Régnier, Achaz, Lambert, Cowie, Bouchet and Fontaine2015). Since the start of the 21st century it has become clear that population depletion and extinction of both freshwater and marine fishes is a severe and widespread problem (e.g. Ricciardi & Rasmussen, Reference Ricciardi and Rasmussen1999; Myers & Worm, Reference Myers and Worm2005; Olden et al., Reference Olden, Hogan and Zanden2007; Burkhead, Reference Burkhead2012).
Extinction of freshwater fishes has been relatively well documented in North America (e.g. Miller et al., Reference Miller, Williams and Williams1989; Burkhead, Reference Burkhead2012). A compilation of the conservation status of freshwater fishes in Mexico has revealed that 10 species have become extinct in the wild or have been extirpated from the country, and > 200 (40% of all species in Mexico) are facing extinction (IUCN, 2016; Ceballos et al., Reference Ceballos, Díaz-Pardo, Martínez-Estévez and Espinosa Pérez2017b; Table 1).
While compiling these data we found that the Catarina pupfish Megupsilon aporus is the most recently extinct endemic freshwater fish in Mexico. This species, which belongs to the Cyprinodontidae and is one of the smallest fishes in North America, was lost as a result of desiccation of the only freshwater spring it inhabited, in El Potosi, Galeana, Nuevo León (Miller et al., Reference Miller, Minckley and Norris2005; Bennett & Conway, Reference Bennett and Conway2010). Unfortunately, this is not an isolated case. Major threats to freshwater fishes in particular and freshwater biodiversity in general have been grouped into five categories: overexploitation of water resources, water pollution, flow modification, destruction or degradation of habitat, and invasive species (Silk & Ciruna, Reference Silk and Ciruna2005; Dudgeon et al., Reference Dudgeon, Arthington, Gessner, Kawabata, Knowler and Lévêque2006). Inland waters in Mexico have experienced these threats, and currently 70% of main rivers are polluted, > 5,000 dams have modified their flow and > 100 invasive species inhabit inland waters (Carabias & Landa, Reference Carabias and Landa2005; Domínguez-Domínguez et al., Reference Domínguez-Domínguez, Boto, Alda, Pérez-Ponce de León and Doadrio2007; CONAGUA, 2011).
The effect of these threats on freshwater fishes depends on the particular environmental conditions, the type and amount of pressure exerted by human activities, and the distribution range of species. For instance, species with restricted distributions, such as the Cuatro Cienegas platyfish Xiphophorus gordoni, which inhabits desert springs, and the Yucatan tetra Astyanax altior, endemic to cenotes in the Yucatan Peninsula, are more imperilled than those with a wide distribution (Jelks et al., Reference Jelks, Walsh, Burkhead, Contreras-Balderas, Díaz-Pardo and Hendrickson2008; Contreras-MacBeath et al., Reference Contreras-MacBeath, Brito Rodríguez, Sorani, Goldspink and McGregor Reid2014). Meanwhile, habitats in northern Mexico with major water supply problems as a result of semi-arid and arid conditions are at higher risk of desiccation than aquatic systems in south-east Mexico (Contreras-Balderas & Lozano-Vilano, Reference Contreras-Balderas and Lozano-Vilano1993; Carabias & Landa, Reference Carabias and Landa2005; Ceballos et al., Reference Ceballos, Ehrlich and Dirzo2017a,Reference Ceballos, Díaz-Pardo, Martínez-Estévez and Espinosa Pérezb). Watershed management in Mexico therefore requires addressing local conservation issues with a regional perspective, to ensure the maintenance of these habitats and their fish populations in the long term (Stiassny & Bianco, Reference Stiassny, Bianco, Cracaft and Grifo1998; Ceballos et al., Reference Ceballos, Díaz-Pardo, Martínez-Estévez and Espinosa Pérez2017b).
The Catarina pupfish (Plate 1) was discovered in 1961. It was described in 1972 as the only species of the genus Megupsilon, and only 22 years after its scientific discovery it became extinct in the wild. We present a historical synopsis of the species and the causes of its decline, and discuss the implications of the extinction process for the conservation of freshwater fish diversity in Mexico.
The discovery and description of a new species
Available information on the biology and ecology of the Catarina pupfish is scarce but examination of peer-reviewed and grey literature provided us with sufficient details to document the sequence of events and factors that led to the species' decline (Miller & Walters, Reference Miller and Walters1972; Contreras-Balderas & Lozano-Vilano, Reference Contreras-Balderas and Lozano-Vilano1996a,Reference Contreras-Balderas and Lozano-Vilanob; Contreras-Balderas et al., Reference Contreras-Balderas, Almada-Villela, Lozano-Vilano and García-Ramírez2003; Echelle et al., Reference Echelle, Carson, Echelle, Van Den Bussche, Dowling and Meyer2005; Jelks, et al., Reference Jelks, Walsh, Burkhead, Contreras-Balderas, Díaz-Pardo and Hendrickson2008; Bennett & Conway, Reference Bennett and Conway2010; Burkhead, Reference Burkhead2012; Liu & Echelle, Reference Liu and Echelle2013).
The first recorded individual of the Catarina pupfish was collected from the wild by R.R. Miller and H.L. Huddle in 1961, but the species was not described until 1972 (Miller & Walters, Reference Miller and Walters1972). Megupsilon aporus is an endemic Mexican species (Miller, Reference Miller1956). The generic name and species were assigned based on two particular characters: Megupsilon in reference to the huge Y chromosome in males, and aporus in reference to the lack of pores in the cephalic sensory system (Miller & Walters, Reference Miller and Walters1972). Although the species is related to Cyprinodon, the characters that classified it as a separate species, along with those previously mentioned, were a different number of chromosomes between males (47) and females (48), the presence of side scales between the dorsal and anal fins, the lack of a black terminal border in the anal fin of nuptial males, the presence of an anal fin as large as the dorsal fin in females, and particular behavioural traits (see below; Uyeno & Miller, Reference Uyeno and Miller1971; Miller & Walters, Reference Miller and Walters1972).
The small Catarina pupfish exhibited sexual dimorphism. Males were smaller (26 mm in length) than females (36 mm; Miller & Walters, Reference Miller and Walters1972). Adult males were steel blue on the back and sides of the body, with a golden sheen in the caudal peduncle, the caudal fin was orange, and a there was a vertical black bar above and below the pupil on the eyes. Adult females had a golden olivaceous colouration over the body (Miller & Walters, Reference Miller and Walters1972).
The Catarina pupfish was endemic to a freshwater spring at El Potosi, on the west side of Sierra Madre Oriental, and within the Sandia basin, Galeana, Nuevo León, Mexico (24°51′N 100°19′W; Fig. 1). This location, at 1,880 m, was also home to the Potosi pupfish Cyprinodon alvarezi and the freshwater crayfish Cambarellus alvarezi (Rodríguez-Almaraz & Campos, Reference Rodríguez-Almaraz and Campos1994), both also endemic to this small spring. The spring was a remnant of a larger water body in an endorheic basin known as La Hediondilla (Miller & Walters, Reference Miller and Walters1972). It is believed that during the Pleistocene the larger lake was connected with the Rio Conchos basin and Rio Bravo tributaries in the north.
Habitat descriptions by Miller & Walters (Reference Miller and Walters1972) mentioned the presence of a main pond (c. 1 ha) with an inner stream and some secondary ponds. Clear water was present year-round, with depths of 0.4–4 m (Miller & Walters, Reference Miller and Walters1972). These conditions remained until the early 1980s (Plate 2; Rodriguez-Almaraz & Campos, Reference Rodríguez-Almaraz and Campos1994). The vegetation in the area included pennywort Hydrocotyle sp., herbaceous flowering plants Nasturtium sp., water primrose Lugwigia sp., hornwort Ceratophyllum demersum, pondweed Potamogeton sp., duckweed Lemna sp., green algae, and grasses restricted to shallow areas (Miller & Walters, Reference Miller and Walters1972; Guzmán-Cedillo, Reference Guzmán-Cedillo1981).
Natural history and ecology
Megupsilon aporus was considered to be part of a relict fauna isolated from the Rio Grande system more than 5 million years ago (Echelle et al., Reference Echelle, Carson, Echelle, Van Den Bussche, Dowling and Meyer2005). It diverged from the common Cyprinodon ancestor to constitute a monotypic genus in the late Miocene (Echelle et al., Reference Echelle, Carson, Echelle, Van Den Bussche, Dowling and Meyer2005). Following its discovery, several expeditions were made to gather information on the species and its habitat (Table 2). On an expedition in 1968 Contreras-Balderas and Lozano-Vilano found that the flow in the pond was constant year-round and that aquatic vegetation was abundant. They also described the abundance of M. aporus; c. 6,000 individuals were found in pockets of the aquatic vegetation mat, which was mainly composed of Ceratophyllum sp. (Contreras-Balderas & Lozano-Vilano, Reference Contreras-Balderas and Lozano-Vilano1996a). By 1974, individuals of largemouth bass Micropterus salmoides were collected in El Potosi and the abundance of Megupsilon was significantly lower. There is no information regarding the introduction of this exotic fish but its presence was one of the causes of the decline of the Catarina pupfish, although not the most pervasive (Contreras-Balderas & Lozano-Vilano, Reference Contreras-Balderas and Lozano-Vilano1996a; see below).
The Catarina pupfish was oviparous, with no breeding territoriality. The species presented two characteristic behaviours: opercular rotation and jaw-nudging during courtship (Liu & Echelle, Reference Liu and Echelle2013). Courtship and mating took a unique form, with no aggressive behaviour; instead attractive movements were performed to bring about successful spawning. Mating was usually carried out at dawn or dusk and there was no cannibalistic egg consumption nor aggression of any kind among males and females (V. González, pers. comm.).
According to Miller & Walters (Reference Miller and Walters1972), M. aporus coexisted with the Potosi pupfish via niche partitioning. The Catarina pupfish preferred shallow areas close to the surface, whereas the Potosi pupfish occurred predominantly in the deep waters of the spring. The two species had different dietary preferences; the Catarina pupfish was carnivorous, feeding mainly on insect larvae and chironomids, whereas the Potosi pupfish fed mainly on algae.
The habitat, described by Miller & Walters (Reference Miller and Walters1972), remained in good condition for more than 15 years (Rodríguez-Almaraz & Campos, Reference Rodríguez-Almaraz and Campos1994). During 1968–1983 the area covered by the spring contracted and expanded as a result of water extraction for agriculture and subsequent recovery during the wet season. However, in 1985 the spring was reduced to 10% of its former size as a result of intensified groundwater pumping. By 1986 only a shallow irrigation ditch and small tributaries remained (Contreras-Balderas & Lozano-Vilano, Reference Contreras-Balderas and Lozano-Vilano1996a; Plate 2).
Overexploitation of water was the main cause of the decline of M. aporus (Contreras-Balderas & Lozano-Vilano, Reference Contreras-Balderas and Lozano-Vilano1996a). By 1990 more than 80 wells of > 100 m depth had been dug for irrigation of corn and potato fields in El Potosi and Sandia Valleys, causing the desiccation of springs and creeks in the region. As a result, aquatic species, including the fishes Cyprinodon veronicae, Cyprinodon longidorsalis, Cyprinodon ceciliae and Cyprinodon inmemoriam became locally extinct (Contreras-Balderas, Reference Contreras-Balderas, Minckley and Deacon1991). The Catarina pupfish was considered to be extinct in the wild by 1994, and in 1995 El Potosi spring was completely dry (Tveteraas, Reference Tveteraas1993; Contreras-Balderas & Lozano-Vilano, Reference Contreras-Balderas and Lozano-Vilano1996b).
Given the grave situation, a strategic plan to protect the area was presented by the first Mexican fish recovery team, which comprised researchers from Universidad Autonoma de Nuevo León. However, the proposal was not adopted or enforced (Contreras-Balderas, Reference Contreras-Balderas, Minckley and Deacon1991).
In the early 2000s natural coal deposits in the area started burning as a result of the collapse and subsidence of sediments, causing the ground to sink in both the El Potosi and La Sandia valleys. Burning episodes have been intermittent, with wet and dry periods (Valdés González, Reference Valdés González1997; Amezcua Torres, Reference Amezcua Torres2009).
Captivity and extinction
One of the first attempts to protect the Catarina pupfish was made by Salvador Contreras-Balderas, who in 1974 moved numerous individuals to other springs in the area to reduce the pressure exerted by the largemouth bass (Contreras-Balderas & Lozano-Vilano, Reference Contreras-Balderas and Lozano-Vilano1996a). One year later, a small population of M. aporus was identified in a small spring in the area. In 1976 Armando J. Contreras-Balderas and his team eradicated the largemouth bass, removing > 280 individuals from the spring (Contreras-Balderas & Lozano-Vilano, Reference Contreras-Balderas and Lozano-Vilano1996a). In the following years, the rapid degradation of the habitat as a result of water extraction increased the need to maintain the species in captivity to avoid extinction. Since 1984 the Catarina pupfish remained in captivity in a management unit specifically for the conservation of the species.
In the early 1990s the New York Aquarium and Universidad Autonoma de Nuevo León established a programme (Emerging rescue of Mexican fish species in imminent danger of extinction) to protect the most representative and threatened freshwater fishes of Mexico. The effort was also joined by the Children's Aquarium at Fair Park, Dallas, Texas. The aims of the programme were to sustain stocks of species over the long term, to maintain sufficient individuals to repopulate native habitats, and to exchange specimens to ensure the diversity of the genetic pools, all by recruiting museums and public aquariums with sufficient space and technical and economic resources.
Individuals of the Catarina pupfish were maintained in various places, including San Luis Potosi in Mexico, Denver and Arkansas in the USA, Zaragoza in Spain, and Germany. However, the species was particularly sensitive to environmental conditions and prone to infections. In 2011 most individuals of the captive populations succumbed to Mycobacteria infections and those that remained were too weak or sterile.
In 2012 the Children's Aquarium at Fair Park, Dallas, Texas, held the last surviving population, with only 20 individuals (Liu & Echelle, Reference Liu and Echelle2013). The last two males were used in an attempt to save the species through hybridization and backcross with a female of C. alvarezi. The reproduction was successful, yielding more than 30 newborn individuals; however, all of them were females with mycobacteriosis, and all eventually died.
The extinction of the Catarina pupfish has implications at various scales. Megupsilon aporus was a monotypic genus that diverged from Cyprinodon c. 7 million years ago according to molecular analyses (Echelle et al., Reference Echelle, Carson, Echelle, Van Den Bussche, Dowling and Meyer2005). The loss of this species precludes any opportunity to study its unique evolutionary history and to understand the adaptive processes by which it was able to survive as a relic for thousands of years. Its extinction, as suggested by Burkhead (Reference Burkhead2012) for freshwater fishes in general, is evidence of a much larger problem: the current loss of freshwater biodiversity.
Freshwater fishes are being lost at a faster rate than terrestrial species. Jelks et al. (Reference Jelks, Walsh, Burkhead, Contreras-Balderas, Díaz-Pardo and Hendrickson2008) reported 700 imperilled freshwater fishes in North America; of the 280 species categorized as Endangered, 65 are endemic to Mexico (Jelks et al. Reference Jelks, Walsh, Burkhead, Contreras-Balderas, Díaz-Pardo and Hendrickson2008; Burkhead, Reference Burkhead2012). In the 20th century Mexico has lost 16 species, 15 of them endemic (IUCN, 2016; Ceballos et al., Reference Ceballos, Ehrlich and Dirzo2017a,Reference Ceballos, Díaz-Pardo, Martínez-Estévez and Espinosa Pérezb; Table 3). The extinction of these species is attributed to two main causes: habitat degradation and pollution. Habitat degradation by desiccation of water bodies affected species that inhabited springs in arid and semi-arid environments, which are under water stress as a result of low precipitation and a high demand for water for human activities (De la Vega, Reference De la Vega2003). Species that were affected by pollution lived in environments close to large cities or to industries that discharge pollutants to the water without regulation or previous treatment (Alcocer, Reference Alcocer2007). The species that became extinct were vulnerable, microendemic species whose life histories were adapted to isolated environments and particular conditions. However, the increasing pressure on freshwater resources may cause a generalized extinction trend, affecting even widely distributed species.
Concern for Mexican freshwater fishes has led to an increase in knowledge about this group and the habitats on which they depend. For instance, in 2012 a national-level analysis of the gaps in the conservation of aquatic biodiversity in Mexico was published (Aguilar et al., Reference Aguilar, Alarcón, González, Kolb, Koleff and Urquiza-Haas2012). This effort, which involved more than 250 specialists, resulted in the identification of priority sites and their representativeness in the national system of protected areas. Although the results indicated that only 15.8% of the area covered by priority sites is currently protected, it was a first step to increase action focused on the conservation, management and restoration of aquatic systems in Mexico.
In 2016 Mexico included 17 freshwater fishes as part of the Alliance for Zero Extinction, an international initiative to safeguard places with microendemic species (AZE, 2018). Local efforts to protect the species and their habitats have increased awareness among the population. For instance, the designation of the RAMSAR site Manantiales geotermales de Julimes, in 2014, was supported by the presence of the Julimes pupfish Cyprinodon julimes. This killifish was described in 2009 and its distribution is limited to El Pandeño hot spring in Chihuahua, Mexico (De la Maza-Benignos & Vela-Valladares, Reference De la Maza-Benignos, Vela-Valladares and De la Maza-Benignos2009). The local community, together with non-profit organizations, researchers and the local government have developed a conservation strategy for the area, which includes the establishment of a state park (Blando-Navarrete et al., Reference Blando-Navarrete, Jiménez-González, Valencia-Castro, Castañeda-Gaytán and Carrillo-Flores2007; De la Maza-Benignos, Reference De la Maza-Benignos2009).
Research conducted by Mexican academic institutions, sometimes in collaboration with international partners, has increased the ex situ conservation efforts for imperilled freshwater fishes. Universidad Michoacana de San Nicolas de Hidalgo and Universidad Autonoma de Nuevo León have led the efforts, with more than 40 species kept in captivity (Lascuráin et al., Reference Lascuráin, List, Barraza, Díaz Pardo, Gual Sill and Maunder2009). In particular, the golden skiffia Skiffia francesae and tequila splitfin Zoogoneticus tequila are examples of species that otherwise would be extinct, as they no longer exist in the wild (Lascuráin et al., Reference Lascuráin, List, Barraza, Díaz Pardo, Gual Sill and Maunder2009).
There is still an urgent need to implement programmes focused on the maintenance of freshwater systems. Lack of regulation and enforcement are threatening both freshwater fishes and water provision. The extinction of the Catarina pupfish is a reminder of the vulnerability of aquatic systems. There is a need to improve ecosystem management strategies and the policies related to freshwater resources, and the development of a national strategy for the conservation of freshwater fishes should be a priority. Such a strategy should include reintroduction of species extinct in the wild, securing populations of the most threatened species both in captivity and in the wild, and reduction of threats. This strategy would bring together scientists and practitioners to identify species and priority sites for conservation, design a monitoring system to assess the current state of the populations, improve watershed management practices, and increase the impact of ex situ and in situ conservation actions to maintain species in the long term. A concurrent citizen science programme in which local communities would play a main role in the protection and monitoring of the habitats could have a positive impact, with relatively low investment costs. There is still time to avert further extinctions of fish species in Mexico, but the window of opportunity is closing. The next 2 decades will be critical for most species at risk.
We thank the Instituto de Ecologia, Universidad Nacional Autónoma de México, for support, and Rosalba Becerra for preparing the figure.
Conceptualization: GC; compilation of information on the Catarina pupfish: AVG, MEAV, LME; writing: LME, GC.
Conflicts of interest
This research complied with the Oryx Code of Conduct for authors.