Providing evidence to improve practice

Action: Restore wetland Amphibian Conservation

Key messages

  • Seventeen studies investigated the effectiveness of wetland restoration for amphibians.
  • Ten site comparison studies (including eight replicated studies) in Canada and the USA compared amphibian numbers in restored and natural wetlands. Eight found that amphibian abundance, numbers of species and species composition were similar. Two found that the number of species or abundance was lower and species composition different in restored wetlands. One found that restored wetlands were used more or less depending on the habitat surrounding natural wetlands. One global review found that in 89% of cases, restored and created wetlands had similar or higher amphibian abundance or numbers of species to natural wetlands.
  • Seven of nine studies (including six site comparison and/or replicated studies) in Canada, Taiwan and the USA found that wetland restoration increased numbers of amphibian species, with breeding populations establishing in some cases. Three found that numbers of species or abundance did not increase with restoration. Two found mixed effects, with restoration maintaining or increasing abundance of individual species. Three replicated studies (including two site comparison studies) in the USA found that numbers of species in restored wetlands were affected by wetland size, proximity to source ponds and seasonality, but not wetland age.
  • Three studies (including two replicated, site comparison studies) in Taiwan and the USA found that restored wetlands were colonized by three to eight amphibian species. One before-and-after study in the USA found that three target species did not recolonize restored wetlands.


Supporting evidence from individual studies


A before-and-after, site comparison study in 1995–1996 of a degraded forested wetland in South Carolina, USA (Barton et al. 2000, Bowers et al. 2000) found that restoration increased numbers of amphibian species over the first four years. Sixteen frog and toad and 13 salamander species were captured in the restoration area. It was assumed that there were no amphibians prior to restoration. Successful reproduction was documented for 16 of the 29 species. However, species diversity was lower in the restored compared to natural site. Planting regimes and treatment (burning or herbicide application) had little effect on species assemblage. Restoration included tree planting in 1993–1995 (549–1078 trees/ha). In some areas herbicide application and prescribed burns were undertaken to control scrub. Approximately 25% of the restoration area was left as unmanaged (control) strips. Amphibians were monitored over 21 months in planted and unplanted areas and in adjacent natural wetland area.



A before-and-after, site comparison study in 1995–1998 of a wetland restoration site in St. Clair County, Illinois, USA (Mierzwa 2000) found that by the end of the study, all seven species of amphibians previously found at the site were recorded within the restored area. An eighth species, not present in the adjacent forest, had also colonized the site by 1997. Abundance was higher at the restored site compared to the adjacent forest (5 vs 4 amphibians/man-hour of survey). Restoration of the 95 ha area included removal of low embankments to restore water levels and planting native hardwood trees. Amphibians were monitored at the restoration area and an adjacent forest in May–June 1995–1996. Drift-fences (15 m long) with pitfall and funnel traps at the centre and ends were used. In 1997–1998, visual encounter surveys were carried out twice in March–May.



A replicated, site comparison study in 1999–2000 of 13 wetlands in South Dakota, USA (Juni & Berry 2001) found that combined amphibian and reptile species richness did not differ significantly between restored, enhanced and natural wetlands. Although not significant, there was a trend for higher numbers of species in restored and enhanced wetlands compared to natural wetlands. A total of 11 amphibian and reptile species were recorded. Study sites were four restored, four enhanced and five natural wetlands. Restoration tended to involve plugging drainage ditches or breaking sub-surface drainage tiles. Enhancement included manipulating water levels to increase wetland size or changing vegetation structure. Monitoring was undertaken using drift-fences with pitfall traps and visual surveys around wetland perimeters in spring and autumn in 1999–2000.



A replicated, site comparison study in 1998 of seven restored wetlands in Minnesota, USA (Lehtinen & Galatowitsch 2001) found that eight amphibian species rapidly colonized the wetlands and four of those established breeding populations. Natural wetlands supported an additional four species. However, there was no significant difference between average numbers of species in restored and natural wetlands (4 vs 5). Six of the seven restored wetlands supported amphibian populations. Species richness increased with restored wetland size and proximity to source ponds. Wetlands were restored 5–20 months before the study, by destroying drainage tiles or filling ditches to allow flooding. Five natural wetlands were surveyed for comparison. Wetlands were seasonal to semi-permanent and were 0.1–8.6 ha in size. Amphibians were monitored on five visits in April–July 1998 using visual encounter and call surveys along wetland edges. Larval sampling was also carried out using five activity and five minnow traps along pond edges.



A replicated, site comparison study in 1999–2000 of 97 restored wetlands in aspen parkland in Alberta and Saskatchewan, Canada (Paszkowski, Puchniak & Gray 2001) found that restored wetlands had similar amphibian abundance to natural wetlands, suggesting that restored wetlands provide suitable habitat for amphibians. A total of 4,086 wood frogs Rana sylvatica, boreal chorus frogs Pseudacris maculata and tiger salamanders Ambystoma tigrinim were captured over the two years. Amphibians, in particular wood frogs, were found in similar numbers in restored and natural wetlands. From 1987, wetlands were restored by installing ditch plugs. Amphibian presence/absence was recorded in 97 restored and 85 natural wetlands. In 1999, seven and in 2000, 11 restored and natural wetlands were monitored intensively. Pitfall (19,431 trap nights) and minnow traps (6,794 trap nights) were used to compare species between restored and natural ponds.



A replicated, controlled study in 1998–1999 of 22 restored wetlands on Prince Edward Island, Canada (Stevens, Diamond & Gabor 2002) found that restored wetlands had significantly higher numbers of amphibian species than non-restored wetlands (2.7 vs 1.8). All five species present on the islands were recorded in both wetland types. Abundance was significantly higher in restored wetlands for spring peeper Pseudacris crucifer (2.7 vs 2.1), northern leopard frog Rana pipiens (0.5 vs 0.2) and green frog Rana clamitans (0.7 vs 0.2). There was no difference in abundance of wood frog Rana sylvatica or American toad Bufo americanus in restored and non-restored wetlands. Wetlands were 0.3–0.6 ha and had been restored by dredging (30–95% of area) two to seven years before the study. Amphibians were monitored at 22 dredged and 24 undredged wetlands during monthly call surveys in May-July 1998 and/or 1999.



A replicated, site comparison study in 1999 of 11 restored and 29 abandoned wetlands on old mines in southwestern Indiana, USA (Timm & Meretsky 2004) found that species richness was similar at reclaimed and abandoned wetlands. Restored wetlands supported an average of 3–8 species and abandoned wetlands 4–6 species. Two natural wetlands supported eight of nine local species. Breeding was recorded at both wetland types. The emphasis of reclamation was restoring mined lands to the original land use (e.g. forestry and agriculture). However, standards being followed included actions that enhanced reclaimed wetlands by developing the shoreline and establishing ephemeral wetlands. Wetlands included permanent (average 11 ha), semi-permanent (3 ha) and ephemeral sites (<0.3 ha). Call surveys were undertaken over three hours in February–August 2000. Tadpole surveys were conducted in March–August using dip-nets, minnow traps and seines. Twenty-nine abandoned and two natural wetlands were used as comparisons.



A before-and-after study in 1998–2003 of a wetland landscape restoration project at Kankakee Sands, Indiana, USA (Brodman et al. 2006) found that numbers of amphibian breeding populations increased from 14 to 172 and species richness from seven to 10, three years after restoration began. Prior to restoration in 1998, there were 14 populations of seven species at seven breeding sites (> 200 m apart). By 2000, this increased to 33 populations of seven species at 14 sites and by 2003, 172 populations of 10 species at 44 sites. Average species richness/site increased from two in 1998 to four in 2003. Species became significantly more common and breeding occurred in every land management unit (vs 50% in 1998). However, apart from in wetter than average years (2002 and 2003), restored wetlands dried before larvae of most species metamorphosis. Restoration began in 1999 and comprised plugging and filling ditches, breaking drainage tiles and recontouring basins. Amphibians were monitored at restored and natural wetlands in April–July 1998, 2000–2003. Call surveys (3/year), seines, dip-nets, minnow traps (each 2/year), terrestrial searches and drift-fences with funnel traps were used.



A replicated, site comparison study in 2003–2004 of two restored wetlands in southwestern Washington, USA (Henning & Schirato 2006) found that amphibian species richness was similar and abundance tended to be higher in restored compared to natural wetlands. Abundances were significantly higher at restored and one natural wetland compared to the other three natural wetlands. Restored and natural wetlands had similar species richness (4–5 species). Pacific treefrogs Pseudacris regilla were only found in natural wetlands. Abundances of the other five species varied between wetlands. Significantly higher number of amphibians emigrated from the restored compared to natural oxbow wetlands (29–58 vs 0.01–0.25/trap night). Abundance was highest in wetlands with intermediate hydroperiods (>7 months) compared to those with temporary or permanent water. Two restored (emergent) and four natural (emergent and oxbow) wetlands were surveyed. Restoration in 1997–1998 involved blocking drainage ditches by constructing water control structures and embankments. Amphibians were monitored using fyke nets and one-way traps in January–June 2003–2004.



A before-and-after study in 2002–2003 of a restored wetland in a tropical forest in Kenting, Taiwan (Lee et al. 2006) found that eight of 18 amphibian species known to be in the area colonized the wetland within a year. A total of 1,456 amphibians were recorded (average density: 0.025 m2). Cricket frog Fejervarya limnocharis was the most common species (62%), followed by ornate narrow-mouthed frog Microhyla ornate and spot-legged treefrog Polypedates megacephalus. These three species accounted for 97% of the relative frequency and abundance. Abundance varied with habitat type and within ponds was positively correlated with vegetation cover. From December 2002 to April 2003, a concrete pond was demolished, the hole filled with soil and replanted to restore a 0.5 ha semi-natural permanent wetland. Amphibians were monitored by visual survey within six habitat areas. Surveys were undertaken twice a month from May 2003 to April 2004.



A small, replicated study in 1999–2004 of three seasonal ponds and 200 potholes created at a forested wetland restoration site on Sears Island, Maine, USA (Vasconcelos & Calhoun 2006) found that wood frogs Rana sylvatica and spotted salamanders Ambystoma maculatum colonized and reproduced in the three ponds and bred in 28% of potholes. Spring peeper Pseudacris crucifer colonized and bred in one pond and American toad Bufo americanus visited but did not breed in two ponds. Reproductive success varied between ponds for wood frogs (0.2–48.4 juveniles/egg mass) and spotted salamanders (1.8–5.4). Metamorphosis of these species was only completed in one pothole before drying. In 1997, two ponds were excavated within the original wetland (350 and 600 m2) and one dry detention basin was converted to a pond (900 m2). Approximately 200 small potholes (0.3–110 m2) were also created. Amphibians were monitored in the three ponds in March–October using enclosure drift-fencing. Pitfall traps were installed in pairs every 10 m either side of fences (9–18 pairs/pond). Eggs were counted within ponds and 50 potholes.



A replicated, site comparison study in 2004–2005 of 14 restored and eight created wetlands associated with hardwood forests in Louisiana, USA (Barlow 2007) found no significant difference in amphibian species richness between restored/created and natural wetlands (2.9 vs 2.9). Twelve of 13 species in the area were found within the wetlands, one of which was only found in restored wetlands (upland chorus frog Pseudacris feriarum). Species richness was higher in 2004 (restored: 3.7; natural: 4.2) than 2005 (restored: 2.4; natural: 2.2). Species richness was positively associated with water depth, canopy cover, flooding, aquatic vegetation and surrounding forest. Temporary and permanent wetlands were 1–174 ha and had been restored 1–18 years previously. Restoration had included replanting trees, water management and dredging. Eight natural wetlands within a wildlife refuge were used for comparison. Amphibians were monitored by call surveys (two/season), egg mass counts (one/season) and dip-netting (monthly along 100 m transect).



A replicated study in 1992–2004 of 16 restored wetlands in Wisconsin, USA (Nedland, Wolf & Reed 2007) found that amphibian communities stayed the same as wetlands matured over 12 years. The six amphibian species and overall amphibian abundance did not change between 1992 and 2004 (13–14 calls/wetland). Overall wetland species and coefficients of conservation values increased over time (coefficient: 3.6 vs 3.9). Restoration occurred between 1988 and 1991. Amphibians were monitored at eight wetlands. Four amphibian call surveys were undertaken at each in April–July. Plants and birds were also monitored.



A before-and-after study in 2000–2006 of a restored forested wetland in Lake County, Illinois, USA (Sacerdote 2009) found that restoration did not increase amphibian species diversity or natural recolonization by three target species five years after restoration. There was no natural recolonization by spotted salamander Ambystoma maculatum, wood frog Lithobates sylvaticus or spring peeper Pseudacris crucifer. Species richness was similar before (4–8) and after restoration (4–6); the diversity index tended to increase (0.5 vs 1.2). Post-restoration, the abundance of northern leopard frog Lithobates pipiens, American toad Anaxyrus americanus and western chorus frog Pseudacris triseriata increased. Green frogs Lithobates clamitans and bullfrogs Lithobates catesbeiana were detected in small numbers, but did not breed. Restoration was undertaken in 2000. Agricultural drainage tiles were removed to restore water levels to previous wetland levels. Non-native European buckthorn Rhamnus cathartica and garlic mustard Alliaria petiolaris were removed using herbicide, chainsaws, manually and controlled burns. Native trees were also planted. Amphibians were monitored in 2004–2006 using drift-fences with pitfall traps and funnel traps, dip-netting, artificial cover, visual and mark-recapture surveys.



A replicated, controlled, site comparison study in 2008 of 18 sites within a large wetland restoration area in Florida, USA (Dixon 2011) found that restored sites had higher amphibian species richness and abundance than non-restored sites. Species richness was significantly higher in restored compared to non-restored (8 vs 5–6) wetlands and similar to natural (7–8) wetlands. Abundance was significantly higher in restored compared to non-restored wetlands one year (27 vs 10) but not four years after restoration (18 vs 17). Abundance was highest in natural wetlands (28–35). Species assemblages differed between wetland types. Overall, five species of tadpole were found in restored wetlands (mainly 1/wetland) compared to none in non-restored sites. Natural wetlands contained six species (mainly 1/wetland). Restoration within a failed residential development site involved plugging canal systems to restore previous water levels. Permanent ponds and ephemeral wetlands were created. There were two areas, restored either one or four years previously, each with three replicates of three wetland types: restored, non-restored and natural. Amphibians were monitored during six call surveys in May–September and monthly dip-netting in June–August 2008.



A replicated, site comparison study in 2005–2006 of four restored wetlands in restored grasslands in the Prairie Pothole Region, USA (Balas, Euliss Jr. & Mushet 2012) found that the restored wetlands were used more frequently by amphibians than wetlands within farmland, but not as much as natural wetlands within native prairie grasslands. This was true for two frog, one toad and one salamander species. Four wetlands from each category were selected: farmed (drained with ditches), conservation grasslands (wetland hydrology restored, area reseeded with perennial grassland ≤ 10 years previously) and native prairie grasslands (natural). Call surveys, aquatic funnel traps and visual encounter surveys were undertaken biweekly in May–June 2005–2006.



A review in 2012 of studies examining restored and created wetlands across the world (Brown et al. 2012) found that amphibian species richness or abundance at restored and created wetlands tended to be similar or greater than at natural wetlands. Species richness or abundance of some or all species was greater at restored or created wetlands in 54% of studies, similar in 35% of studies and lower than natural wetlands in 11%. Restored and created wetlands tended to be larger, deeper and were wet for more of the year than natural wetlands. Species richness and abundance tended to be positively associated with abundance of emergent vegetation, proximity of source wetlands and the availability of wetlands with varying water levels. They were also influenced by upland habitat and tended to be negatively associated with fish presence. Only peer-reviewed studies were included (n = 37; 70% in USA). Only studies that converted existing upland or shallow-water areas to wetland habitat (created; n = 27), or restored wetlands (n = 14) were included. Wetlands built specifically for water quality improvement were not included. Twenty-six studies had controls, either natural reference wetlands or historic data.


Referenced papers

Please cite as:

Smith, R.K., Meredith, H. & Sutherland, W.J. (2017) Amphibian Conservation. Pages 9-65 in: W.J. Sutherland, L.V. Dicks, N. Ockendon & R.K. Smith (eds) What Works in Conservation 2017. Open Book Publishers, Cambridge, UK.