Two subspecies, Chelydra serpentina serpentina (Linnaeus, 1758) and C. s. osceola Stejneger, 1918, were recognized until osceola was synonymized with serpentina by Shaffer et al. (2008), based on lack of significant genetic differentiation. Phillips et al. (1996) removed the two former subspecies rossignonii and acutirostris by elevating these to species level, based on genetic differentiation.

The Snapping Turtle (Chelydra serpentina) was previously evaluated as Least Concern (LC) on the IUCN Red List (van Dijk 2012) and we assess it again as Least Concern at this time. The species has since been listed in CITES Appendix III (2016) with a recent up-listing to CITES Appendix II (2023). Chelydra serpentina remains widespread and relatively abundant throughout its native range in the eastern half of North America. Decades of large-volume trade for consumption and the pet trade has seeded non-native populations in regions of western North America, Europe, and Asia. Chelydra serpentina is a habitat and dietary generalist, capable of living in a variety of aquatic habitats and tolerant across a range of climates and human disturbances. Ecologically common species such as C. serpentina play key roles in freshwater wetland ecosystems, but they also bear the brunt of habitat destruction, ecosystem degradation, and overexploitation. The species is harvested from the wild and farmed in large volumes in the United States and increasingly in China to support markets for consumption, pets, and unsubstantiated medicine (folk medicine, traditional Chinese medicine). Export volumes of live C. serpentina from the United States have been substantial in recent decades, ranging from tens of thousands to hundreds of thousands of known wild-caught individuals annually and totals of hundreds of thousands to greater than one million individuals annually.

Despite their generalist habits and high starting abundance, populations of C. serpentina are limited in their ability to tolerate added mortality. Considering the geographic range and variety of habitats occupied by C. serpentina, populations are likely to experience multiple concurrent threats (e.g., harvest, road mortality, subsidized predators, fishhook ingestion). The cumulative effect of several concurrent threats can easily elevate mortality and gradually cause decreased population abundance. Very high adult survival (greater than 90% per annum) combined with a long and slow life history—characterized by low egg, hatchling, and juvenile survival, slow growth, late age at maturity, low reproductive output spread over a long lifespan, and long generation time—makes populations susceptible to decline. Small increases in adult mortality, especially of reproductive females, can lead to sustained population declines that are difficult to detect. There is strong evidence of population declines caused by acute (e.g., predation) and chronic (e.g., commercial harvest, road mortality) reductions to survival in C. serpentina populations, particularly in the northerly populations in the United States and Canada. Some populations have demonstrated no evidence of recovery even decades after threats were abated, despite adult survival having rebounded to high levels and population connectivity maintained. The absence of population recovery on management-relevant timescales warrants a precautionary approach to conservation and the maintenance of existing populations rather than attempting recovery once populations have already declined. The high starting abundance of C. serpentina means that they have further to decline before conservation action is taken and, of critical importance, further to recover in order to restore populations to some semblance of their pre-impact abundances. The high starting abundance, longevity, and continued persistence at sites of relatively high disturbance can easily mask long-term population declines. The goal of conservation and management for C. serpentina should be to keep this “common” species common.

Chelydra serpentina currently qualifies as Least Concern (LC) status on the IUCN Red List. Moving forward, the careful monitoring of population, harvest, and trade trends is necessary. State and federal agencies in the United States need to collaborate, use harvest and trade data to make evidence-based decisions around harvest/export, and continue to refine regulations around wild harvest. Conservation and management actions should include: implementation of meaningful harvest restrictions at US state levels; addressing uncertainties about the scale of domestic harvest (recreational and commercial) and processing of turtle meat for international markets; enacting regulations to address farm stocking, including closing loopholes that allow for the laundering of wild turtles through farms; and regular review of export quotas in light of monitoring of wild and farm populations of C. serpentina.

The North American Snapping Turtle, Chelydra serpentina, ranges throughout North America east of the Rocky Mountains (Iverson 1992, Ernst and Lovich 2009, TTWG 2021). In Canada, the species is present in the southern portions of the provinces of Saskatchewan, Manitoba, Ontario, Quebec, New Brunswick, and the province of Nova Scotia. Northernmost range records occur at approximately 51°N latitude in Manitoba and Ontario (Holman and Andrews 1994; Manitoba Herp Atlas Project; Ontario Reptile and Amphibian Atlas). In the USA, C. serpentina reaches a natural western range boundary in the states of Montana, Wyoming, Colorado, and New Mexico, and is widespread eastward to the Atlantic coast. The species is native to 41 of the 48 conterminous USA states. Southernmost range records occur at approximately 25–26°N latitude in the southwest approximated by the Nueces River in Texas and in the southeast at the southern tip of peninsular Florida. Naturally occurring individuals of the species have never been confirmed in northern Mexico, but it is plausible that populations may exist (Smith and Smith 1979, Legler and Vogt 2013). The estimated historical indigenous extent of occurrence (EOO) is 9,022,386 sq. km, and the estimated historical indigenous range (area of occupancy, AOO) is 5,250,199 sq. km (TTWG in press).

The species has been introduced in many places, including the following: Canada (British Columbia), Belgium, China (eastern and southeastern China, approximated by Liaoning, Heibei, Shandong, Jiangsu, Anhui, Zhejiang, Hubei, Hunan, Jiangxi, Guangdong; Liu et al. 2021), France, Germany, Hong Kong, Italy (Esposito et al. 2022), Japan (Chubu region of Honshu Island; Kobayashi et al. 2006), Mexico, Republic of  Korea (Koo et al. 2020), Taiwan, Province of China (Shiau et al. 2006), and USA outside its native range (Arizona, California, Nevada, northwestern New Mexico, Oregon, Utah, Washington).

There are many highly localized (e.g., lake or river system) population estimates for Chelydra serpentina available in the literature. No estimates or extrapolations are available at landscape scales, such as provinces, states, or countries. The species is widely and continuously distributed across a variety of habitats in eastern North America. It is rarely the numerically dominant turtle species in any location, but it usually represents about 2–30% of the total number of turtles and a high proportion of turtle biomass in wetland habitats (e.g., Dreslik et al. 2005, Aresco and Gunzburger 2007, Browne and Hecnar 2007). The species presumably has a substantial ecological significance given its large body size, relatively high densities of 1.2–74 turtles per hectare, biomass estimates of 19–340 kg per hectare of suitable habitat (Iverson 1982, Congdon et al. 1986; Galbraith et al. 1988, Congdon and Gibbons 1989, Iverson et al. 2000, Johnston et al. 2012), and large potential for nutrient cycling and nutrient standing stock (e.g., Sterrett et al. 2015, Wenger et al. 2019). The ecological role of common species, such as C. serpentina, is generally understudied and underappreciated; however, even relatively small proportional declines in the abundance of common species can result in large absolute losses of individuals and biomass, causing significant disruptions to ecosystem structure, function, and services (Gaston and Fuller 2008, Gaston 2010, 2011).

On a landscape scale, C. serpentina appears relatively tolerant to habitat loss given its persistence in areas of high wetland destruction, road density, and semi-urban spaces; however, care must be taken when interpreting occupancy and abundance at finer spatial scales (e.g., Paterson et al. 2021), which can be the spatial scales at which key processes such as population connectivity and ecosystem function are occurring. Major threats, such as wetland destruction and road mortality, are likely to negatively affect populations of C. serpentina at fine spatial scales. Increased road networks and density have led to male-biased populations as females make more overland migrations for nesting (Gibbs and Steen 2005). Also, the persistence of at least one animal (occupancy) can mask declines in abundance. The long and slow life history of C. serpentina, very high natural survival of adults, high percentages of egg predation and mortality of hatchlings, and long generation time make the species vulnerable to population declines and slow to recover.

Habitat types: Chelydra serpentina is a habitat generalist (Ernst and Lovich 2009, Paterson et al. 2012) capable of living in a wide variety of aquatic habitats. Still or slow-moving bodies of water with a soft mud or sand bottom and abundant aquatic vegetation are preferred habitats (Ernst and Lovich 2009). Typical aquatic habitats include bogs, creeks, fens, lakes, marshes, permanent or temporary ponds, rivers, and swamps. Human-modified water bodies, such as agricultural ponds and stormwater retention ponds, are also used, as are brackish coastal habitats. Aquatic habitats used in areas of the introduced range are similar to those in the native range (e.g., Kobayashi et al. 2006). Habitat use can differ between summer and winter, with the flooded root systems in riparian woodlots serving as hibernacula (Ryan et al. 2014). In the northern range, aquatic dormancy and overwintering occur amongst submerged debris or in the soft bottom of waterbodies (Ultsch 2006). Chelydra serpentina can undertake long-distance aquatic and terrestrial migration (greater than 10 km round trip) to access nest sites (Obbard and Brooks 1980). Females nest in open-canopy habitats with high sun exposure, such as in forest clearings, meadows, shorelines, rock outcrops, beaches, agricultural fields, and road shoulders. Nest sites range from less than a meter to a few hundred metres from a water body. The species will shift habitat use away from the preferred habitat of Alligator Snapping Turtles (Macrochelys temminckii) when the two species occur in microsympatry (Riedle et al. 2015, Lescher et al. 2013). 

Chelydra serpentina is a generalist omnivore (Ernst and Lovich 2009). Plant matter can be a major component of the diet (Alexander 1943; Lagler 1940, 1943; Aresco and Gunzburger 2007; Lewis and Iverson 2018). Animal prey items include amphibians, fishes, crayfish, molluscs, aquatic insects, carrion, and, infrequently, birds and mammals.

Males can be combative and sustain serious injury from male-male sexual competition (Keevil et al. 2017). The mating system might be one of convenience polyandry in which females acquiesce to the mating attempts of much larger males (Keevil et al. 2017). Mate searching, combat, and copulatory behaviour are concentrated in spring but can occur throughout the active season (Brown and Brooks 1993). Clutches demonstrate multiple paternity (Galbraith et al. 1993).

The estimated generation length for C. serpentina is 25 years (Michigan, Congdon et al. 1994) to 31 years (south-central Ontario, COSEWIC 2008). Average sizes for females and males (straight midline carapace length, SCL): female 28–32 cm, male 35 cm (Keevil et al. 2017, Hughes and Meshaka 2020). Median male C. serpentina are 1.2 times longer in SCL and 1.8 times heavier than females (Ontario, Canada; Keevil et al. 2017). Maximum sizes for females and males (straight midline carapace length): female 39.0 cm, male 50.3 cm (Gerholdt and Oldfield 1987, Pritchard 1989, Iverson et al. 1997, Johnston et al. 2012, Walde et al. 2016). The size at hatching (straight midline carapace length, SCL) has been recorded as 27–29 mm in Ontario (Riley et al. 2014); 16–38 mm (mean 28.8 mm) in Pennsylvania (Ernst and Lovich 2009); and 35 mm SCL in Florida (Aresco et al. 2006).

Clutch size is positively correlated with latitude. Mean clutch sizes are approximately 35–51 eggs (up to approximately 100 eggs in exceptional cases) in northerly and westerly areas of the species’ geographic range and approximately 20 eggs in southerly areas (Iverson et al. 1997). See Iverson et al. (1997) for clutch size data across the geographic range of C. serpentina.

At constant incubation temperatures spanning 22–30°C, full-term embryonic development takes 55–95 days (Yntema 1978). In the northern range of the United States and in Canada, incubation duration typically exceeds 90 days (Congdon et al. 1987, Riley and Litzgus 2013). In these northerly regions, high rates of annual nest failure can occur because the developmental period is too cool or too short to allow for full-term embryonic development before the onset of lethal freezing temperatures (Obbard and Brooks 1981).

Sexual maturation is related to size rather than age (Armstrong and Brooks 2013). Size and age at maturity vary with latitude, with northerly and westerly populations experiencing slower annual growth and a relatively later onset of size at maturity. In south-central Ontario, female C. serpentina reach sexual maturity at 24 cm SCL at approximately 15–20 years of age (Galbraith et al. 1989; Armstrong and Brooks 2013, 2014). In Nebraska, females mature at 28–29 cm SCL at 10–12 years of age (Iverson et al. 1997). Female age at maturity is 11–16 years in Michigan (Congdon et al. 1994). Other sizes and ages at maturity for female C. serpentina include: 23 cm SCL in Iowa (Christiansen and Burken 1979), 27.5 cm SCL and 12 years in Texas (Suriyamongkol et al. 2023), and 18.5–20.2 cm SCL and 6–8 years in Florida (Aresco et al. 2006, Aresco and Gunzburger 2007). Following sexual maturity, females reproduce approximately annually (0.85–1.0 reproductive frequency per year) up to a maximum of one clutch per year (Congdon et al. 1994, 2008; Iverson et al. 1997; Keevil et al. 2021). In Michigan, the annual fecundity of reproductive females (considered as the number of daughters a reproductive female can produce) was estimated as 12 eggs, based on a mean clutch size of 28 eggs, a reproductive frequency of 0.85, and an estimated half of all eggs producing females (Congdon et al. 1994). Reportedly, farmed C. serpentina in China are capable of nesting three times per year (Shi et al. 2004).

Annual survival of adult C. serpentina is naturally very high: 97–98% in Ontario (Galbraith and Brooks 1987. Brooks et al. 1991, Armstrong et al. 2018, Keevil et al. 2018), 93–96% in Wisconsin and Minnesota (Paisley et al. 2009), 93–97% in Michigan (Congdon et al. 1994), 90–95% in Pennsylvania (Hughes and Meshaka 2020), 97% in West Virginia (Flaherty et al. 2008), 91% in Virginia (Colteaux and Johnson 2017b), 91–99% in North Carolina (Eskew et al. 2010), 95% in Nebraska (Iverson 2023), and 93–94% in Texas (Rose and Small 2014). Females with reproductive histories spanning at least 40 years are known from long-term study in Ontario, Canada (Keevil et al. 2018). Individuals are confirmed to live to at least 70 years (COSEWIC 2008) and life span estimates greater than 100 years are plausible (Armstrong and Brooks 2014).

Chelydra serpentina is widely exploited, in some areas at the northern limits of occurrence to the degree that populations have declined significantly. As a result, Canada and many states in the U.S. have brought the species under various degrees of regulation or legislation to manage exploitation and trade, with a regulated or unrestricted commercial offtake permitted in about half the U.S. states in which the species occurs. Collection of Snapping Turtles from the wild, and captive production in turtle farms, for export to East Asia increased substantially in recent years, from about 10,000 animals declared as exported from the USA in 1999 to over 300,000 annually in recent years (LEMIS database - total recorded export numbers: 1999 – 10,053; 2000 – 18,486; 2001 – 38,911; 2002 – 63,644; 2003 – 129,683; 2004 – 141,544; 2005 – 316,500; 2006 – 377,408; 2007 – 316,093; 2008 – 558,491; available data do not allow differentiation of farmed vs. wild-collected trade volumes).

While habitat loss and degradation likely affect individuals and populations, Snapping Turtles are adaptable and mobile and are less likely than many other turtle species to be threatened by habitat change processes. The impact of 'subsidized' predators (i.e., unnaturally large populations of predators subsidized by easily available resources near human settlements) is generally unquantified but may be significant. The impact of pollution, particularly of pesticides, industrial chemicals, and heavy metals, on Snapping Turtles appears to be sub-lethal at worst, to the point that snappers are a recognized biomonitoring species to monitor the accumulation of pollutants in aquatic ecosystems. Road-kill and other human-induced mortality occurs, but it is not clear whether this represents an overall threat to the species in combination with other impacts.

Vulnerability due to slow life history.—Like turtles generally, the long and slow life history of C. serpentina makes the species vulnerable to decline. Small increases in adult mortality will lead to steady population declines. Harvest pressure of as little as 10% of adults over 15 years of age can halve population size in as little as 15 years (Congdon et al. 1994). Populations are most sensitive to elevated mortality of adults, especially adult females (Doak et al. 1994, Cunnington and Brooks 1996, Enneson and Litzgus 2008). Long generation times (25–30 years, or more) make population recovery very slow following declines, if recovery occurs at all. Literature review did not uncover any examples of C. serpentina populations with documented recovery following decline (population recovery, especially in the absence of an intensive management programme, has been rarely reported among turtle populations). A three-year mortality event that caused annual survival to temporarily decline from 94% to 76–86% in a C. serpentina population reduced the population by half and there has been no evidence of recovery in the following 22 years (Brooks et al. 1991, Keevil et al. 2018). This lack of recovery at management-relevant timescales and the risk of further declines strongly support prioritizing the protection of populations rather than relying on recovery after declines have occurred (Keevil et al. 2018). Similar findings are reported for other turtle species in North America (e.g., Pitt and Nickerson 2013, King et al. 2016). Regional populations of C. serpentina are known to have experienced substantive declines that have resulted in conservation assessments and status listings, particularly in Canada. The high starting abundances, the longevity of individuals, and continued presence at sites (albeit at possibly declining abundances and densities) can easily mask long-term population declines.

Harvest, domestic and international markets.—Chelydra serpentina is exploited as food, in the international pet trade, and for unsubstantiated medical use (folk medicine, traditional Chinese medicine). The species is significantly less abundant in regions with commercial harvest compared to non-harvested areas (Shaffer et al. 2017, Colteaux and Johnson 2017b), supporting previous findings that even legal harvest regulations in U.S. states allow unsustainable levels of exploitation (Zimmer-Shaffer et al. 2014). All harvest estimates presented throughout this account are to be considered underestimates because of incomplete and conservatively estimated data.

Habitat degradation, fragmentation, and destruction.—There has been widespread degradation, fragmentation, and destruction of primary habitat for C. serpentina across the eastern half of North America over the past two centuries (equivalent to approximately six to eight generations of the species). For example, in southern Ontario, the centre of the distribution of C. serpentina in Canada, more than 70% of wetlands were drained over the past 200 years (Glooschenko 1983, Snell 1987, Ducks Unlimited Canada 2010). This region has the most thermally favourable and productive wetland conditions for C. serpentina in Canada and hence these populations have (or had) the highest natural abundances and densities in the region (COSEWIC 2008). It can be reasonably inferred that the destruction of most wetlands in southern Ontario has resulted in large-scale population declines of C. serpentina in the region and in the Canadian population at large (See Appendix 1 of COSEWIC [2016] for an assessment of the decline in the relative abundance of an Ontario turtle species attributable to historic wetland conversion). Similar logic extends to many regions of the central and eastern United States with histories of large-scale wetland destruction and urbanization. States throughout the range of C. serpentina have experienced major wetland destruction from the 1780s to 1980s (equivalent to approximately 6–8 generations of C. serpentina): Alabama (-50% wetland coverage), Arkansas (-72%), Colorado (-50%), Connecticut (-74%), Delaware (-54%), Florida (-46%), Georgia (-23%), Illinois (-85%), Indiana (-87%), Iowa (-89%), Kansas (-48%), Kentucky (-81%), Louisiana (-46%), Maine (-20%), Maryland (-73%), Massachusetts (-59%), Michigan (-50%), Minnesota (-42%), Mississippi (-59%), Missouri (-87%), Montana (-27%), Nebraska (-35%), New Hampshire (-9%), New Jersey (-39%), New Mexico (-33%), New York (-60%), North Carolina (-49%), North Dakota (-49%), Ohio (-90%), Oklahoma (-67%), Pennsylvania (-56%), Rhode Island (-37%), South Carolina (-27%), South Dakota (-35%), Tennessee (-59%), Texas (-52%), Vermont (-35%), Virginia (-42%), West Virginia (-24%), Wisconsin (-46%), and Wyoming (-38%) (Dahl 1990). Ignoring other threats, wetland destruction alone is almost certainly responsible for massive declines in the relative abundance of C. serpentina over the past one to two centuries in the central and eastern United States. Snapping Turtles can use human-modified aquatic habitats, such as agricultural and wastewater treatment ponds. However, it is important to consider that the presence/occupancy of C. serpentina in degraded, low-quality or otherwise marginal habitats is likely not ecologically equivalent to abundances in minimally impacted, high-quality habitats. The human-modified habitats make C. serpentina more susceptible to contaminants due to water quality and stormwater runoff (Ryan et al. 2013).

Road and railway mortality.—The first documented account of road-killed wildlife is reportedly that of C. serpentina crushed under the wheel of a loaded wagon in North Dakota in 1897 (Knutson 1987). There is an extensive literature reporting the major negative impacts of mortality experienced by turtles on roads, C. serpentina included (e.g., Ashley and Robinson 1996, Haxton 2000, Gibbs and Shriver 2002). The species has a home range of several square kilometres, which increases its risk of road morality or trauma (Ryan et al. 2013). Additionally, increased road networks have led to male-biased populations as females make more overland migrations for nesting (Gibbs and Steen 2005, Steen et al. 2006). Even low levels of additional mortality, especially of reproductive females, can have detrimental effects on the long-term sustainability of turtle populations. Juvenile mortality on roads is largely overlooked but can have negative demographic consequences that are equally severe as adult mortality (Keevil et al. 2023). Similarly, a largely unquantified threat is the entrapment and mortality of turtles along extensive railway networks in North America.

Subsidized predators.—The impact of subsidized predators (i.e., unnaturally large populations of predators subsidized by easily available resources near human settlements; e.g., Raccoons that depredate Snapping Turtle nests) demonstrates major regional variation generally unquantified and can be a major cause of low nest success. In Algonquin Provincial Park, Ontario, an area largely free from subsidized predators, 19% of Snapping Turtle nests were destroyed by predators (Obbard and Brooks 1981). Subsequent research in Algonquin Provincial Park estimated nest survival as approximately 0.40–0.70 (Riley and Litzgus 2014). Nest survivorship ranged from 0 to 0.70, with a long-term average of 0.23 in the population at the ES George Reserve, Michigan (Congdon et al. 1987, 1994). In human-disturbed habitats of Point Pelee National Park, Ontario, nest survivorship averaged 0.16 over two years, with the high depredation rate attributable to subsidized Raccoons (Wirsing et al. 2012). Other reports of nest depredation rates include 59% in South Dakota (Hammer 1969) and 94% in northern New York state (Petokas and Alexander 1980).

Environmental pollution.—Snapping Turtles are relatively tolerant of high levels of pollution, including pesticides, industrial chemicals, and heavy metals, and have been used as a biomonitoring species to monitor pollutants in aquatic ecosystems (e.g., de Solla et al. 2001, 2008). Environmental contamination is known to feminize male C. serpentina (de Solla et al. 1998), increase embryonic deformities (Bishop et al. 1998), reduce hatching success (Bishop et al. 1991, 1998; Golet and Haines 2001; de Solla et al. 2008), and might have long-term population consequences (Rowe 2008). The longevity, omnivorous diet, and high trophic position of C. serpentina tends to result in high levels of contaminants in their tissues through bioaccumulation, bioconcentration, and biomagnification. Tissue of C. serpentina can readily exceed levels of mercury, polychlorobiphenyls (PCBs), and persistent organic pollutants considered safe for human consumption (Hebert et al. 1993, Golet and Haines 2001, Sherwood 2017). Other physical and chemical forms of environmental pollution, such as siltation and eutrophication, have unknown consequences for C. serpentina.

Miscellaneous causes of mortality by humans.—Chelydra serpentina is often vilified and targeted for human persecution because of its large size, fearsome appearance, and unjust reputation for aggression and as a voracious predator of waterfowl and sport fish. The species is vulnerable to direct persecution by humans in many forms: intentional road mortality (Ashley et al. 2007), blunt trauma and recreational firearms target practice (“plinking”; Shook et al. 2023), boat and propeller strikes (including in protected areas that allow motorized boat traffic; Bennett and Litzgus 2014, Smith et al. 2018), lead poisoning and ingestion of hooks from recreational and sport fishing (Borkowski 1997, Brown and Sleeman 2002, Steen et al. 2014, Steen and Robinson 2017), and bycatch mortality in freshwater fisheries (Raby et al. 2014, Midwood et al. 2014, Shook et al. 2023). The severity of these threats remains largely unquantified but recent research suggests they are underestimated and of concern when combined with other known and pervasive threats (e.g., habitat destruction, road mortality, commercial harvest).

Infectious disease.—Several diseases have been reported from populations of C. serpentina and sympatric turtle species in recent years. Ranavirus infections vary from low to high prevalence in wild populations (Carstairs 2019, Carstairs et al. 2020), and are known to cause mortality of C. serpentina (McKenzie et al. 2019). Herpesviruses and Mycoplasma sp. have been detected in C. serpentina without apparent ill effects, but infections by these agents are known to cause illness and mortality in other chelonian species (Alplasca et al. 2019).

Climate change.—Chelydra serpentina exhibits temperature-dependent sex determination with females produced at low and high temperatures and males produced at intermediate temperatures. Climatic warming, as is generally expected across the native range of C. serpentina, could potentially shift the production of sex ratios (O'Steen 1998). However, in a population near the northern climatic range limit, no shift in hatchling sex ratio was reported across 48 years despite significant warming of the environment during this period (Leivesley et al. 2022). The resilience of populations to climate change remains to be investigated elsewhere in the species' range.

Chelydra serpentina is heavily exploited for food, pets, and unsubstantiated medicinal use (folk medicine, traditional Chinese medicine). Consumption ranges from local, subsistence collection to commercial trade for local, national, and international markets. Specimens available in the market can be of wild-caught origin, commercially farmed within the native range (United States), or commercially farmed outside the native range (especially China). Wild-caught individuals can also supply the commercial farming sector inside and outside the native range. Chelydra serpentina is among the most highly harvested, farmed, and exported turtle species in the United States (Reed and Gibbons 2003, Mali et al. 2014).

The international demand for turtles from North America coincides with the severe overexploitation and collapse of turtle populations in Asia. China is the chief market for the consumption, medicinal use, and trade of turtles (van Dijk et al. 2000, Luiselli et al. 2016). In North America, principally the United States, C. serpentina is subject to exploitation because of its large body size (high meat yield), broad geographic distribution, and relative abundance. Chelydra serpentina is one of the most common species available for harvest and consumption in the United States, particularly in the Mid-Atlantic states where a sizable harvest, processing, and meat canning industry exists (Cain et al. 2017) and in the southeastern states where a culture of turtle consumption persists (Roman and Bowen 2000) and export numbers are high (Mali et al. 2014).

There is no legal commercial or recreational harvest of C. serpentina in Canada. In the United States, the recreational and commercial harvest (regulated and unregulated) of C. serpentina is still permitted in selected states. Conservatively estimated, the harvest of C. serpentina in the USA increased 209% from 1998 to 2014 (Colteaux and Johnson 2017a). As of 2015, “19 of the 37 states” within the native range of C. serpentina in the USA were open to commercial harvest (Colteaux and Johnson 2017a; note that the present assessment reports C. serpentina as native to 41 states of the United States). Of the 19 states allowing commercial harvest at that time, eight had a minimum size limit and five had limits on the number of turtles that could be harvested over a given time period (Colteaux and Johnson 2017a). At least Connecticut, South Carolina, and Texas have eliminated commercial harvest of Snapping Turtles since 2017. See Mali et al. (2014), Colteaux and Johnson (2017a), Sherwood (2017), and Bennett and Loda (2020) for a summary of commercial turtle trapping laws of the United States. Requirements and regulations for recreational (personal) and commercial harvest are highly variable between states. In some jurisdictions, only a standard fishing license, sometimes with a nominal surcharge, is required to harvest C. serpentina, whereas other states maintain a monitored commercial permit system (Cain et al. 2017, Sherwood 2017). Several states maintain weak or non-existent laws and poor practices regulating the commercial overexploitation of turtles, including unlimited harvest allowances, nondisclosure of harvest data, absent or voluntary harvest reporting requirements, poor regulation of commercial permits, and/or a lack of season, size limit, or geographic restrictions on turtle harvest (see Mali et al. 2014, Sherwood 2017, Bennett and Loda 2020).

Overall, from 1998 through 2013 (16 years), a minimum estimated 348,529 C. serpentina were reported as commercially harvested among the 11 US states for which harvest data were available (Colteaux and Johnson 2017a). The commercial harvest of C. serpentina is particularly active in some Mid-Atlantic states (Maryland, Delaware, New Jersey, Virginia, North Carolina; Thorbjarnarson et al. 2000, Cain et al. 2017, Colteaux and Johnson 2017a, Sherwood 2017). In the United States, the wholesale purchase, processing, and resale of wild-harvested C. serpentina is seemingly concentrated in the Mid-Atlantic states, notably Maryland. During the period 2008–2016, more than 70,000 C. serpentina (equivalent to more than 1,000,000 pounds, or 454,000 kg, of live turtles) were harvested in Maryland (Maryland Department of Natural Resources 2016). Annually during 2008 and 2016, 19 to 55 harvesters reported wild harvests of approximately 74,000 to 144,000 pounds (33,600 to 65,500 kg) of C. serpentina, with the top three harvesters in each year accounting for 34% to 69% of the annual harvest (Maryland Department of Natural Resources 2016). A major turtle wholesaler in Maryland purchased 4,500 to 9,000 kg of C. serpentina per day in May and June, equating to the capture of 1,000 to 2,000 average-sized adult female turtles per day (but note that both sexes were harvested; Cain et al. 2017). Harvesters travel from other states, some quite distant (e.g., Ohio, Connecticut), to sell their harvest in Maryland (Cain et al. 2017). In New Jersey, the number of registered commercial harvesters and the number of harvested C. serpentina reported in the period 2014–2022 have gradually declined significantly as compared to the higher levels recorded in 2009–2013 (B. Zarate pers. comm.).

Records from Virginia indicate that 721,000 pounds (327,700 kg) of C. serpentina were commercially harvested from 2002 to 2015 (Colteaux and Johnson 2017a). Annual harvest ranged from approximately 11,500 to 125,500 pounds (5,200 to 57,000 kg) or, on average, 3,100 individuals (up to nearly 8,000 individuals) annually (Colteaux and Johnson 2017b). These values are underestimates because there is an indication that not all harvesters are fully reporting their harvests and private harvest is not reported or monitored (Colteaux and Johnson 2017b). In Virginia, harvest pressure was estimated to cause a 17% reduction in adult survival in harvest areas (Colteaux and Johnson 2017b), which exceeds values known to cause steady population declines (Congdon et al. 1994) and are equivalent to reductions in survival that have resulted in failed population recovery on a quarter-century timescale for C. serpentina (Keevil et al. 2018). In Virginia, densities of C. serpentina are 47% to 62% lower at harvest sites compared to an unharvested site (Colteaux and Johnson 2017b). Similarly, in Missouri, abundances of C. serpentina were drastically reduced in harvested locations (mean of 15 turtles) compared to unharvested locations (mean of 90 turtles; Shaffer et al. 2017). Such harvests are not sustainable. Chelydra serpentina has been the most harvested species in Iowa since 1987 (Fowler 2020, cited in Dolan 2020) and the contemporary absence of larger individuals as well as regional declines in catch-per-unit-effort are highly suggestive of a negative effect of harvest on wild populations (Dolan 2020).

Several states in the United States have recently restricted or eliminated commercial harvesting of freshwater turtles and/or Snapping Turtles (e.g., Texas in 2007 and 2018, Florida in 2009, Alabama in 2012, Iowa in 2017, Connecticut in 2018, South Carolina in 2020; Mali et al. 2014, Bennett and Loda 2020, M.J. Ravesi pers. comm.) or enacted regulations on harvest (e.g., minimum size limits, daily catch limits, and/or closed seasons and closed areas for C. serpentina; Cain et al. 2017, Colteaux and Johnson 2017a). Tightening regulations and enforcement in one jurisdiction can shift harvest efforts and trafficking to neighbouring states with weaker regulations (e.g., see Mali et al. 2014, Colteaux and Johnson 2017b). Virginia is one such example, having maintained lax regulations while nearby states (Maryland and North Carolina) enacted stricter harvest policies. Over 15 years (2002–2017), the number of commercial turtle harvesting permits sold to out-of-state persons in Virginia increased from 1 to 26 and harvest of C. serpentina increased by 1,300% (Colteaux and Johnson 2017b). From 2009 to 2013, out-of-state harvesters, representing only a quarter of all permit holders, accounted for as much as 70% of the annual reported harvest in Virginia (Colteaux and Johnson 2017b). A similar scenario might explain an increase in turtle exports, including C. serpentina, from Louisiana following commercial harvest closures in Alabama and Florida (Mali et al. 2014).

Collection of Snapping Turtles from the wild and captive production in U.S. turtle farms for export to East Asia increased substantially over the past two decades. Based on data from the United States Fish and Wildlife Service Law Enforcement Management Information System (LEMIS), exports from the USA have increased approximately linearly from 7,000 turtles declared as exported from the USA in 1999 to over 1,324,000 exported annually in the late 2010s (Colteaux and Johnson 2017a). From 1999 to 2011, the export of wild-harvested C. serpentina fluctuated between approximately 18,000 and 68,000 individuals per year, but exports in 2012 and 2014 increased an order of magnitude to 249,609 and 207,383, exported wild individuals, respectively (Colteaux and Johnson 2017a). In general, the proportion of wild-caught turtles supplying export has fluctuated from 8% to 24% (Colteaux and Johnson 2017a).

Export data underestimate the number of wild-harvested C. serpentina in two ways (Colteaux and Johnson 2017a). Firstly, an unknown (unreported) number of individuals/biomass are processed and canned as meat in the USA before export. Reporting exports of processed Snapping Turtle meat is required to the U.S. Fish and Wildlife Service (B. Weissgold pers. comm.); however, a lack of records in the Law Enforcement Management Information System of the U.S. Fish and Wildlife Service make clear that there is extensive non-compliance in reporting by exporters. The lack of trade data for canned meat and soup exports of C. serpentina is inconsistent with the large processing volumes and large overseas markets. Overall, the lack of reported trade data for C. serpentina is concerning and indicates a failure of traders’ adherence to regulations. Secondly, the distinction between wild and farm-produced stock is questionable because there can be fraudulent classification of wild-caught turtles as exported as farmed supply after having passed through farms. That is, it is unclear whether farms are truly supplying export demand through captive breeding or to what degree they are restocking with wild captures and laundering wild-caught turtles into export markets.

From 2002 to 2012, approximately 100,000 C. serpentina were exported from Texas, 125,000 from Florida, 665,000 from Louisiana, and 4,349,000 from California (Mali et al. 2014). The number of exported C. serpentina of wild-caught origin significantly increased from 2002 to 2012, notably among exports from California (Mali et al. 2014). Notably, California received C. serpentina from elsewhere in the United States given that the species is not native to the state, any within-state harvest of introduced populations would not come close to supporting such an export quantity, and there is no large-scale farming of the species in the state.

The trade data in Table S1 (see Supplementary Information) involving C. serpentina were retrieved from CITES Wildlife Tradeview (2023), as reported by exporters from 2016 to 2021: total trade amounts to 831,886 live individuals and only 84 eggs. Table S1 provides source data for traded C. serpentina. Export and import summaries by nation: USA exported 831,620; Mexico exported 350; China imported 719,847; Hong Kong imported 93,049; Macau imported 18,090; Spain imported 500; France imported 300; UK imported 100, and Canada imported 84.

The trade data in Table S2 (see Supplementary Information) involving C. serpentina were retrieved from CITES Wildlife Tradeview (2023), as reported by importers for the period of 2016 to 2021: total trade amounts to 647,472 live individuals, 260,000 specimens, and 1 carapace; USA exported 907,462; China exported 9; and Canada exported 2. Table S2 provides source data for traded C. serpentina. Export and import summaries by nation: China imported 797,765; Hong Kong imported 103,593; Indonesia imported 5,400; Spain imported 500; Netherlands imported 200; USA imported 11; Belgium imported 3; and Australia imported 1.

It is not clear what accounts for the many discrepancies between CITES trade data for C. serpentina as reported by exporters and importers. Discrepancies can arise for many reasons, such as traders shipping smaller numbers than their permit allows (e.g., exporters may report permit quantity, whereas importers may report actual quantity), as well as an export (application) date and actual export occurring in different calendar years. Discrepancies between reported data from importers and exporters include: countries of export, countries of import, the origin of traded individuals (notably among the categories “wild” and “captive-bred”), year-specific volumes of trade (especially in 2020 and 2021), total volumes of trade from 2016 to 2022 (a difference of over 75,000 units), and the purported state of units (notably between the categories “live” individuals and “specimens”).

A 2008 survey of recognized turtle farms in China generated an estimated captive population size of 32,000 C. serpentina with annual sales of 4,100 individuals (based on a 45% survey response rate; Shi et al. 2008). More recently, Liu et al. (2021) estimated the online sale of more than 100,000 snapping turtles (Chelydra spp. and Macrochelys spp. combined) through a single e-commerce platform in China. However, much of China’s production of C. serpentina involves the rearing of hatchlings (aquaculture) acquired from elsewhere. Summing exports (hatchlings) from elsewhere and apparent production in China is not necessarily accurate because it risks double-counting the same turtles. Farming efforts and populations of C. serpentina in international farms are likely grossly underestimated.

Cain et al. (2017) reported that the harvest pressure and market value of C. serpentina as meat in the eastern United States is highly dependent upon market forces, such as demand and domestic sales vs. international sales. Depending on market demands, wild harvested females are exported live for aquaculture (with gravid females particularly sought after; Colteaux and Johnson 2017b), whereas males tend to be killed, processed, and their meat canned for export (Cain et al. 2017, Colteaux and Johnson 2017a). A harvester was estimated to generate a gross income of $14,000 to $28,000 USD for a 31-day trapping season, with an estimated daily harvest of 204 kg at $2.20 to $4.41 USD per kilogram of C. serpentina (Cain et al. 2017). Sherwood (2017) reported the value of wild harvested C. serpentina in New Jersey ranged from approximately $0.65 to $2.00 USD per pound ($1.43 to $4.40 per kg) when sold to meat processing factories or seafood vendors and $1.00 to $2.50 USD per pound ($2.20 to $5.50 USD per kg) when sold to restaurants. Females were preferred and generated higher prices ($0.75 to $1.00 USD more per pound, equivalent to $1.65 to $2.20 USD per kilogram, for female turtles; Sherwood 2017). In New Jersey, surveyed harvesters reported earning up to $10,000 per year from harvest sales (Sherwood 2017).

Ceballos and Fitzgerald (2004) reported values of $3.00 to $10.89 USD per kg for turtle meat sold in Texas (turtle species unknown, but presumably C. serpentina and/or Apalone spp.). Online markets in the United States sell meat for $44 to $66 USD per kilogram of unknown turtle species (Mali et al. 2015). The value of farmed-raised C. serpentina in China was approximately $25 USD per individual or $45 USD per kilogram of meat (Shi et al. 2008).

In the U.S. pet trade, individual juvenile C. serpentina are available for as little as $6 USD (Montague et al. 2022). Online platforms selling undersized C. serpentina violate U.S. federal regulations prohibiting the sale of turtles less than 4 inches (10 cm) in size (Montague et al. 2022).

Chelydra serpentina is listed as a species of Special Concern by the Committee on the Status of Endangered Species in Canada (COSEWIC 2008) and the federal government of Canada. This designation recognizes the species as particularly sensitive to human activities or natural events; that it has declined to a level of abundance at which its persistence is increasingly threatened by genetic, demographic or environmental stochasticity, but the decline is not yet sufficient to qualify the wildlife species as Threatened; and that the species may become Threatened if factors suspected of negatively influencing the persistence of the wildlife species are neither reversed nor managed with demonstrable effectiveness. Chelydra serpentina has been listed in CITES Appendix II since 2023; previously (since 2016) it was listed in Appendix III (USA).

The species does not have a federal conservation status in the United States, but occurs in a substantial number of protected areas, both public and private. Conservation designations vary at the state level. According to NatureServe, this species is Secure (S5) in North America. Available long-term trend data suggest that populations appear stable, but some have increased in area as they colonized ponds created by humans. Chelydra serpentina is considered a Secure species because it is widespread in North America and occupies disturbed habitats and colonizes new habitats. Most natural heritage databases rank this species as Secure (S5) and Apparently Secure (S4). In Montana in the United States, and in Saskatchewan, Manitoba, New Brunswick, and Nova Scotia in Canada, it is ranked as Vulnerable (S3). 

In Canada no commercial harvest is allowed in any province. Chelydra serpentina was delisted as a game reptile in Ontario (2017), which closed the recreational hunt of the species. Federal conservation listing of C. serpentina as a species of Special Concern has been in effect for 15 years (COSEWIC 2008). This listing does not confer any habitat protection or conservation actions, but it recognizes past declines and the potential for future declines if threats are not managed. 

In the United States state-level regulations are increasingly addressing the commercial harvesting of turtles; many states have closed harvests of C. serpentina and other species completely. Regulatory strategies vary by state but have included: caps on maximum harvest, size limits, harvest reporting requirements, regulation of commercial harvesting permits/licenses, spatial and/or temporal restrictions on turtle harvest. See Mali et al. (2014), Colteaux and Johnson (2017a), Sherwood (2017), and Bennett and Loda (2020) for the descriptions of turtle harvesting regulations by state.

In both Canada and the United States the goal of conservation and management for C. serpentina should be to keep this “common” species common.

Conservation measures needed:

• Implementation of state-level data collection addressing the number of turtles harvested and butchered domestically for international sale as processed products
• Implementation of state-level commercial permit/licensing system with mandatory harvest reporting requirements
• Implementation of regulations at the state level to address the stocking of farms with wild stock, and the laundering of turtles through farming operations
• Federal coordination among states to collect robust harvest and export data across the range of C. serpentina
• Continued data collection about import and export numbers of C. serpentina        
• Action from the USA to address uncertainties about reporting wild vs. captive (farm) origin of exported C. serpentina, as well as reporting the types of exports (hatchling vs. adult live exports, processed meat, soup, other parts, etc.)
• Action by wildlife inspection authorities to ensure that exports of live animals, meat, soup, and other parts and products of C. serpentina are adequately inspected and reported.