Conservation Evidence strives to be as useful to conservationists as possible. Please take our survey to help the team improve our resource.

Providing evidence to improve practice

Action: Head-start amphibians for release Amphibian Conservation

Key messages

Read our guidance on Key messages before continuing

  • Twenty-two studies head-started amphibians from eggs and monitored them after release.
  • Six of 10 studies (including five replicated studies) in Denmark, Spain, the UK and USA and a global review found that released head-started tadpoles, metamorphs or juveniles established breeding frog populations or increased populations of frogs or toads. Two found mixed results with breeding populations established in 12 of 17 studies reviewed or at two of four sites. Two found that head-started metamorphs or adults did not prevent a frog population decline or establish a breeding toad population. For five of the studies, release of captive-bred individuals, translocation or habitat management were also carried out.
  • Nine of 10 studies (including nine replicated studies) in Australia, Canada, Europe and the USA found that head-started amphibians released as tadpoles, metamorphs or adults metamorphosed successfully, tended to survive the first season, winter or year or bred successfully. One found adult survival was 1–17% over four years and one found limited breeding following the release of adults.
  • Four replicated studies in Australia, the UK and USA found that frog survival to metamorphosis and size at metamorphosis was greater and time to metamorphosis shorter in head-started compared to wild animals. One replicated study in Canada found that young head-started leopard frogs were smaller than those in the wild. One replicated study in Australia found that corroboree frog tadpoles released earlier had higher survival, but metamorphosed two weeks later than those released a month later.
  • Three studies (including one replicated study) in the USA only provided results for head-starting in captivity. Two found that Houston toad eggs could be captive-reared to tadpoles, but only one successfully reared adults. Three studies (including two replicated studies) in Canada and the USA found that during head-starting, amphibian growth rate, size, stress levels and survival was affected by the amount of protein provided, housing density or enclosure location. One found that mass, stress levels and survival were not affected by the amount of food or habitat complexity.


Supporting evidence from individual studies


A replicated study in 1978–1979 of Houston toads Bufo houstonensis at Houston Zoo, USA (Quinn 1980) found that tadpoles were raised successfully from eggs, but problems were encountered raising adults. Mortality rates of tadpoles were 5–9%. Several experimental groups under different conditions demonstrated that toads raised in naturally planted outdoor enclosures grew faster and had significantly higher survival rates than those raised indoors.



A replicated, before-and-after study in 1985–1987 of 20 restored and created ponds near Aarhus, Denmark (Skriver 1988) found evidence of breeding by European tree frogs Hyla arborea a year after head-started metamorphs were released. In 1986, 17–21 males were heard calling in four ponds, but no females, eggs or tadpoles were recorded. In 1987, up to 50 males were heard calling in 13 ponds. Four egg masses were found in one pond and tadpoles in six ponds. One hundred and fifty egg masses were collected from a local wild population. Animals were captive-reared in hot houses. Over 6,000 metamorphs were released into nine created and 11 restored ponds over 10 km2 in 1985–1986.



A before-and-after study in 1972–1991 of natterjack toads Bufo calamita on heathland in Hampshire, UK (Banks, Beebee & Denton 1993) found that captive-rearing and releasing toadlets, along with aquatic and terrestrial habitat management, tripled the population (see also Buckley & Beebee 2004). Egg string counts, i.e. the female population increased from 15 to 43, with a maximum 48 in 1989. Captive-reared toadlets raised from eggs were released in 1975 (8,800), 1979, 1980 and 1981 (1,000/year). Nine small ponds were created (< 1,000 m2) and four restored by excavation. In addition, scrub, bracken and swamp stonecrop Crassula helmsii were removed and limestone was added to one acidic pond annually in 1983–1989. Toads were monitored each 10 days in March and August each year.



A replicated, before-and-after study in 1991–1993 of created ponds on restored opencast mining land in England, UK (Bray 1994) found that released head-started great crested newt Triturus cristatus tadpoles returned as adults and bred in the second year. Adults returned to at least five of eight ponds and larvae were caught in three of five ponds netted in 1993 (2–5 tadpoles/pond). Newt eggs were collected and reared to tadpoles in aquaria. In 1991, 630 tadpoles were released into four ponds and in 1992, 1,366 tadpoles into eight ponds (66–243/pond). Ponds were surveyed using a dip-net in July 1993. Sixteen ponds (30 x 20 m) with shelved edges and terrestrial habitat had been created on restored land. Ponds were planted with submerged and edge plants. Terrestrial habitat created included scrub/woodland, rough grassland, ditches and hedgerows.



A before-and-after study in 1986–1997 of restored and created ponds at six sites in Funen County, Denmark (Briggs 1997) found that releasing head-started toadlets increased the population of European fire-bellied toads Bombina bombina over 10 years. The total adult population increased from 82 in 1986–1988 to 542 in 1995–1997 (from 1–30 to 8–170 toads/site). Numbers of ponds occupied by adults increased from eight to 62 and by tadpoles from one to 18. The population only declined at one site that was flooded with salt water. Wild-caught toads were paired in separate nest cages in ponds and eggs collected and reared in aquaria. Metamorphs and one-year-olds were released into 69 restored and created ponds. Each year, ponds were monitored for calling males and breeding success (capture-recapture estimate) in 1987–1997.



A before-and-after study in 1987–1994 of ponds on Jersey, UK (Gibson & Freeman 1997) found that releasing head-started agile frogs Rana dalmatina did not prevent a decline in breeding within the population. Over 300 head-started toadlets were released. However, frog activity at the release site decreased over the years and there was no breeding in 1991 or 1994. In 1987–1989 and 1992, eggs were collected in the wild, reared to froglets and released back in the wild.



A replicated, before-and-after study in 1994–1997 of garlic toads Pelobates fuscus in Jutland, Denmark (Jensen 1997) found that released head-started tadpoles established breeding populations in the two restored, but not two created ponds. Forty-three adults were also translocated to one of the restored ponds. Authors considered that the failure of created ponds may have been due to predation, because of the lack of vegetation and introduction of sticklebacks Pungitius pungitius. Four egg strings were laid in captivity and produced over 2,000 tadpoles. One thousand tadpoles were released at different stages before metamorphosis into one restored and one created pond. Two ponds had been restored and two created in 1994–1995. Tadpole and call surveys were undertaken.



A replicated study in 1997 at three sites in the Snowy Mountains, Australia (Hunter et al. 1999) found that southern corroboree frog Pseudophryne corroboree survival from eggs to metamorphosis was significantly higher for captive-reared compared to wild tadpoles (53–70% vs 0–13%). That was the case at two of the three sites, at the third, the same trend was seen for average clutch survivorship (captive: 33%; field: 15%). Tadpoles released earlier had higher survival than those released later. However, late-release tadpoles metamorphosed two weeks before early-release tadpoles. Field-reared tadpoles metamorphosed two weeks later than both. A total of 374 eggs were collected from the wild. Late stage tadpoles were returned to field enclosures within their original pools in two batches one month apart. Survival of field and captive-reared tadpoles was monitored by dip-netting once a fortnight until metamorphosis. Water levels were maintained to avoid pool drying.



A before-and-after study in 1995–1999 of boreal toads Bufo boreas in a National Park in Colorado, USA (Muths, Johnson & Corn 2001) found that captive-reared and released toads did not establish a stable breeding population. Eighteen of the 800 released metamorphs were recorded one week after release, but none in 1997–1999. Unmarked metamorphs were found in 1996–1997. Fifty-six of the adult toads released were recaptured during the first three months, but none were seen in following years. Nine eggs masses were collected from the wild in July 1995. Half of each egg mass was captive reared. In September 1995, 800 captive-reared metamorphs were toe-clipped and released. The site was monitored for the following week, twice monthly in May–June 1996 and then weekly. One hundred toads were reared and released in July 1996. These were monitored on alternate days in July–September and then weekly until November. Toads had been absent from the release site for five years.



A replicated study in 1998–2000 of Italian agile frog Rana latastei in the Lombardy District, Italy (Gentilli et al. 2002) found that tadpoles were raised successfully in captivity and metamorphosed once they were released. In 2000, 1,200 agile frog tadpoles were raised successfully. In 2001, the number raised was 28,000. Animals metamorphosed once released in both years. Eggs were collected and hatched in semi-natural conditions in captivity. Tadpoles with developing hind limbs were released back to their original ponds. In 2000, frogs were released back to two sites and in 2001 to six sites. Half of the tadpoles were translocated to new and restored ponds.



A replicated study in 1999–2002 of northern leopard frogs Rana pipiens in Alberta, Canada (Kendell 2003) found limited evidence of breeding following captive-rearing and release of frogs. At one site, seven released frogs were recaptured, a further three were heard calling and one egg mass was observed. Survival to metamorphosis in captivity was 17–33% each year. Three to six egg masses were collected from the wild each year and reared to froglets in two man-made outdoor ponds. Predation was prevented where possible by exclusion or removal of predators. Between 1999 and 2002, a total of 6,500 captive-reared frogs were tagged and released at three new sites. Surveys were undertaken at one release site in May–July 2002.



A replicated, before-and-after study in 1998–2003 of stripeless tree frogs Hyla meridionalis in Gipuzkoa province, Spain (Rubio & Etxezarreta 2003) found that released, captive-reared juveniles, with captive-bred juveniles and translocated adults, established breeding populations in 11 of 14 created ponds. Metamorphs, breeding behaviour, eggs and well-developed larvae were observed in 11 of the ponds. Froglets were also recorded in some ponds. Translocated adults survived in good numbers and returned to 12 of 14 ponds. Introduced predators, dense vegetation, eutrophication and drying resulted in reduced survival and reproduction in some ponds. A small number of additional ponds were colonized by the species. Thirteen ponds were created and one restored with vegetation planted in 1999–2000. Eggs were collected and reared in captivity in outdoor pools. A total of 871 metamorphs and 19,478 tadpoles were released into eight of the ponds. An additional 5,767 tadpoles were bred in captivity and released and 1,405 adults translocated to the ponds.



A replicated, before-and-after study in 1972–1999 of natterjack toads Bufo calamita at two sites in England, UK (Buckley & Beebee 2004) found that captive-rearing toadlets, along with pond creation and restoration and vegetation clearance, increased populations over 20 years. At one site, the continuation of a study in 1972–1991 (Banks, Beebee & Denton 1993) until 1999 indicated that there was a doubling of the population. Egg string counts (i.e. female population) increased from 15 in 1972 to 32 in 1999, with a maximum number of 48 in 1989. At a second site, where head-starting had been undertaken most years since 1980, egg string counts increased from 1 in 1973 to 8 in 1999, with a maximum number of 29 in 1997. Ponds were created and restored by excavation, scrub and bracken was cleared and captive-reared toadlets raised from eggs and released. Toads were monitored annually.



A replicated study in 2000–2005 at two wetlands in British Columbia, Canada (Adama & Beaucher 2006) found that captive-reared and released northern leopard frog Rana pipiens tadpoles and metamorphs survived over winter and bred successfully. At one site, seven juveniles, three adults and 13 unmarked young of the year were recorded the year after release. At the other site three egg masses and numerous young-of-year were recorded in one area, but no frogs were caught in the second area. In 2005, population estimates for young of the year/site were 1,361 and 3,874 respectively. Wild young were significantly larger than captive-reared young in all but two years (13 vs 8 g). Average survival in captivity was 82%. An increased protein diet resulted in increased size at metamorphosis and decreased time to metamorphosis (reduced 75 days). In 2001–2005, 30,065 hatchlings from 27 egg masses were collected and reared in captivity. In total, 10,147 tadpoles and 14,487 metamorphs were marked and released back to the source population and at two restoration sites. Monitoring was undertaken using visual encounter and call surveys.



A review of the effectiveness of 39 release programmes for head-started or captive-bred amphibians (Griffiths & Pavajeau 2008) found that 14 of 17 programmes that could be assessed were considered successful. Seven species (2 toad; 3 frog; 2 newt) showed evidence of breeding in the wild for multiple generations (high success), five species (3 toad; 2 frog) showed some evidence of breeding (partial success) and two species (1 toad; 1 frog) only showed evidence of survival following release (low success). Three programmes were considered unsuccessful and the outcome was not known for the other 19. Species from 16 countries were involved in these release programmes, with a bias towards temperate countries. Half of the species were classified in the top four highest IUCN threat categories (i.e. vulnerable to extinct in the wild).



A replicated, before-and-after study in 1995–2007 of chiricahua leopard frogs Lithobates chiricahuensis at the Phoenix Zoo, USA (Sprankle 2008) found that head-starting and releasing tadpoles and froglets increased populations. With releases, some populations had recovered enough to produce hundreds of egg masses by 2001. By 2007, the number of ponds where frogs had become or were becoming established had increased four-fold. In captivity, over 90% of egg masses survived to froglets or late stage tadpoles, compared to only about 5% reaching metamorphosis in the wild. Egg masses were collected from the wild from the late 1990s. Between 1995 and 2007, over 7,000 tadpoles and frogs were head-started. Froglets and late stage tadpoles were released back to the wild.



A replicated study in 2008 of Jersey agile frogs Rana dalmatina on Jersey, UK (Jameson 2009) found that survival to metamorphosis was higher for head-started animals than those in the wild (15–22 vs 9–17%). However, those with initial protection in the wild (17%) had similar survival to those head-started. There was no significant difference in survival from release to dispersal of head-started tadpoles released earlier (32–46%) or later (40–46%). However, those released later were larger (0.6–0.7 vs 0.5–0.6 g). Head-started metamorphs were larger than those in the wild (0.5–0.7 vs 0.3–0.5 g). Eleven egg masses were collected and raised in aquaria. Tadpoles (n = 4,468) were marked and released back to two ponds in two groups, 10 days apart. Egg masses left in the wild were either protected in mesh bags for two weeks, or for four weeks in bags followed by protection pens. In June–July, frogs were monitored daily using pitfall traps 3 m apart along drift-fences surrounding ponds.



A replicated study in 2006–2010 in Kosciuszko National Park, Australia (Hunter et al. 2010) found that 1–66% of released captive-reared southern corroboree frogs Pseudophryne corroboree survived. Survival was 1–17% over four years for released adults. Breeding males were recorded at one site in 2008 and 2010 and both sites in 2009. Survivorship from eggs to metamorphosis in artificial pools was 35–66% over two years. Tadpoles and metamorphs tended to be larger in artificial compared to natural pools. Chytrid fungus was detected in one of 11 artificial pools in 2008 and one frog in 2009. In January 2006, 196 four-year-old and 15 five-year-old frogs, largely reared from wild-collected eggs, were marked and released across two sites. Six call surveys were undertaken per site in January 2007–2010.  In April–May 2008–2010, fifty wild-collected eggs were placed in 20 artificial pools (400 L tubs) across four natural bog sites. Tubs had a constant water flow, a layer of pond silt and Sphagnum moss. Survival, size and chytrid infection was assessed just before metamorphosis.



A replicated study in 2010–2012 of white-bellied frogs Geocrinia alba at Perth Zoo, Australia (Bradfield 2011) found that about 70% of head-started frogs released survived for at least a year. Eggs were collected from the wild and reared in captivity for 12 months. A total of 70 frogs were released in 2010 and 31 in 2011 at the same site.



A replicated study in 2007–2012 of Houston toads Anaxyrus houstonensis in Texas, USA (Forstner & Crump 2011, Crump 2012) found that eggs were successfully reared to toads in captivity. In 2007, 35% of juveniles survived in captivity and by 2010 survival had increased to 50–55%. By 2012, approximately 700 toads were held in three captive breeding facilities. In 2007, 500 toads were released, in 2009 it was 4,194 and in 2010, 14,728 were released back into the wild at 10 sites. Thirty-one egg strands were collected from the wild and tadpoles raised in biosecure rooms and four outdoor exclosure tubs.



A replicated study in 2006–2011 of common midwife toads Alytes obstetricans and Iberian frogs Rana iberica in a National Park near Madrid, Spain (Martín-Beyer et al. 2011) found that released head-started midwife toads bred successfully and Iberian frog metamorphs survived their first winter. Fifteen radio-tagged adult midwife toads, two males carrying eggs, one pregnant female and a number of tadpoles were recorded. Mortality of metamorphs during the winter was high. A number of Iberian frogs released the previous year and earlier in the year were located. From 2006, all toad tadpoles found in 250 ponds were collected. Larvae were treated against the chytrid fungus using elevated temperatures (> 21°C) and baths in antifungal drugs (itraconazole). Tadpoles were reared in indoor aquariums, in similar environmental conditions to in the wild. Juveniles and adults were released where they were captured. Frog egg masses and tadpoles were also collected from a stream and head-started in aquariums. Tadpoles, juveniles and adults were released in several streams where fish had been removed by electro-fishing. Animals were monitored twice a week during the summer.



A replicated study in 2008–2009 of Ozark hellbenders Cryptobranchus alleganiensis bishop in Missouri, USA (Bodinof et al. 2012) found that 64% of captive-reared animals survived up to a year following release. Five deaths occurred within 30 days of release, three within 50–92 days and five within 126–369 days. Most hellbenders stayed within a small area (90–94%), with only 7% moving over 20 m/day. Home ranges varied widely in the first 4–7 months after release (51–987 m2), but were significantly smaller during the following six months (11–31 m2). Of those that were known to have established a home range, 69% dispersed less than 50 m from the release point. Overall, 77% had entered their core home range within 21 days of release. Thirty-six hellbenders were captive-reared from eggs and were released back to the two original sites. Animals were radio-tracked from May 2008 to August 2009.



A replicated study in 2010 of spotted salamanders Ambystoma maculatum in the USA (Davis 2012) found that housing larvae at low densities resulted in bigger salamanders, higher survival and lower stress levels, similar to larvae in the wild. At different larval densities there were significant differences in body mass (6/tank: 1.8 g; 12/tank: 1.6 g; 30/tank: 0.9 g), survival (94%; 67%; 33% respectively) and stress levels (white blood cell ratios: 0.4; 1.5; 2.2 respectively). At medium larval densities, increased food or habitat complexity had no significant effect on body mass (food: 1.4 g; environment: 1.7 g), survival (89%; 50% respectively), or stress levels (1.3; 0.7 respectively). Egg masses were collected from the wild. Larvae were reared in three replicates of five treatments: starting densities of six, 12 or 30 larvae/1,000 l tank, increased food (12 larvae/tank with triple the zooplankton) or increased habitat complexity (tank filled with sticks and refugia). All tanks had leaf litter on the bottom. Metamorphs were weighed and blood sampled for stress hormone levels.




A replicated study in 2010–2012 of gopher frogs Lithobates capito in southwest Georgia, USA (Hill 2012) found that some head-started froglets survived once released. In 2012, some froglets released earlier that year were observed and a large adult female that had been released in 2010 was re-captured. Portions of egg masses were collected from one of the remaining breeding sites and transferred to institutions for rearing to metamorphosis. Tadpoles were reared outdoors in large tanks with plant matter from the egg collection site. Tadpoles were offered some supplemental feeding, but largely ate the plants provided. Over 4,300 froglets were marked and released onto restored Nature Conservancy land, which lacked a natural population. In 2012, froglets were released directly into burrows as protection from drought. Monitoring began in summer 2012.



A replicated, before-and-after study of agile frogs Rana dalmatina at two sites on Jersey, UK (Wilkinson & Buckley 2012) found that following the release of head-started metamorphs, breeding increased at both sites. The number of egg clumps increased by approximately 500% and the number of breeding ponds occupied increased compared to five years previously. Tadpoles were held in captivity until metamorphosis and then released at existing, re-profiled and newly created ponds at two sites.


Referenced papers

Please cite as:

Smith, R.K., Meredith, H. & Sutherland, W.J. (2019) Amphibian Conservation. Pages 9-65 in: W.J. Sutherland, L.V. Dicks, N. Ockendon, S.O. Petrovan & R.K. Smith (eds) What Works in Conservation 2019. Open Book Publishers, Cambridge, UK.