Includes seven recognized subspecies: Apalone spinifera spinifera (LeSueur, 1827), A. s. aspera (Agassiz, 1857), A. s. atra (Webb & Legler, 1960) A. s. emoryi (Agassiz, 1857), A. s. guadalupensis (Webb, 1962), and A. s. pallida (Webb, 1962). Apalone spinifera atra has at times in the past been treated as a full species, but is currently considered a subspecies of A. spinifera based on genetic and morphological studies (McGaugh 2008, McGaugh and Janzen 2008, McGaugh et al. 2008, TTWG 2021). The previously recognized subspecies Apalone spinifera hartwegi (Conant & Goin, 1948) is currently considered a synonym of A. s. spinifera, based on its limited morphological and molecular distinctiveness (McGaugh et al. 2008, TTWG 2021).

With the exception of Canada and smaller marginal and disjunct populations in the USA, Apalone spinifera appears to be a widespread and generally abundant species in its core range. Moreover, the species’ nesting requirements appear to be met in most systems, although somewhat restricted in northern marginal populations. The few studies comparing numbers over time have found mainly resilient populations in the face of anthropogenic disturbances, notably in some urban or suburban habitats, although some northern marginal and disjunct populations are more sparse and appear to be declining. Several peripheral populations have become genetically and demographically isolated due to large-scale impoundments, especially in the Missouri River system. Overall, we assess the species A. spinifera as being Least Concern (LC) at present; it was also previously assessed in 2011 as Least Concern (van Dijk 2011).

The subspecies A. s. atra was previously assessed in 1996 as Critically Endangered (CR A1ace+2c; criteria 2.3) (TFTSG 1996), and we maintain the CR assessment at this time, but now as A4ce under criteria 3.1 (in prep.), indicating a c. 80% population decline based on decreased area of occupancy (AOO) and habitat quality and the potential effects of hybridization with A. s. emoryi on an ongoing basis over three generations (c. 75 years) extending from two generations in the past and projected for one generation into the future. The taxon also qualifies as Endangered (EN) under criteria B2ab(ii,iii,iv) for a small AOO with few locations and continued decreases in AOO, habitat quality, and number of locations. The other subspecies of A. spinifera have not previously been separately assessed; we assess them at present as follows: A. s. spinifera, A. s. aspera, and A. s. pallida as Least Concern (LC), and A. s. emoryi and A. s. guadalupensis as Data Deficient (DD).

Although there are threats such as human collection, habitat modification, and subsidized predation, the species appears secure throughout most of its USA range. For the species as a whole, there remain concerns that high nest depredation from subsidized mesopredators (primarily Raccoons) and ongoing harvest could result in declining turtle populations. Peripheral populations, such as those in Canada and Mexico, however, warrant focused attention, particularly in the face of climate change and potential barriers to dispersal between river drainages (e.g., highways and other unsuitable habitats). Disjunct populations, such as the ones of A. s. spinifera in Lake Champlain and the upper Hudson Valley, and isolated populations such as those in Montana and Wyoming, are also in need of ongoing conservation efforts. Specific status evaluations for the subspecies A. s. emoryi and A. s. guadalupensis are highly desirable. As always, long-term monitoring is necessary to guard against the species becoming significantly threatened.

Apalone spinifera is native to southernmost Ontario and Quebec in Canada, northern Mexico from Chihuahua to Tamaulipas, and most of the United States east of the Rocky Mountains. It occurs nearly completely throughout the Mississippi-Missouri-Ohio river systems, as far upstream as Montana and Wyoming, the Great Lakes region, Ottawa, southern St. Lawrence and Hudson systems, the Colorado system, the Rio Grande system, Atlantic drainages from Cape Fear to the Saint Marys River, and Gulf drainages from the Apalachicola to the Rio Soto de Marina (Iverson 1992, Ernst et al. 1994, Ernst and Lovich 2009, Powell et al. 2016, TTWG 2021). The species has been recorded up to an altitude of 1,320 m in the Bighorn River drainage at Thermopolis, Wyoming (W.A. Estes-Zumpf, unpubl. data).

The species has been introduced into Arizona, California, western Colorado, Massachusetts, Nevada, New Jersey, western New Mexico, eastern Pennsylvania, Utah, Virginia, and Washington in the USA, as well as Baja California, western Coahuila, and Sonora in Mexico (TTWG 2021, Meshaka et al. 2022). The introduced populations in eastern Pennsylvania appear to be expanding their geographic range (K. Gipe, unpubl. data).

Apalone spinifera spinifera: Most of the Mississippi-Missouri-Ohio and Great Lakes systems, with several isolated populations scattered around the various Great Lakes and in the Hudson Valley and Lake Champlain.
Apalone spinifera aspera: from southern North Carolina to Mississippi (see Moler 2006 for distribution in Florida).
Apalone spinifera atra: endemic to the Cuatro Ciénegas basin of Coahuila, Mexico (see Cerdá-Ardura et al. 2008).
Apalone spinifera emoryi: Rio Grande – Pecos basin of Texas and New Mexico, USA, and Chihuahua to the Rio Soto de Marina in Tamaulipas, Mexico. Also inhabits the Colorado system of Arizona and California (and adjoining Utah, Nevada, and New Mexico), documented to be introduced (Miller 1946, Iverson 1992, Ernst et al. 1994). Co-occurs with Apalone s. atra in the Cuatro Cienegas basin, where it prefers riverine sections (McGaugh and Janzen 2008; Cerdá-Ardura et al. 2008).
Apalone spinifera guadalupensis: South-central Texas.
Apalone spinifera pallida: Louisiana, northeastern Texas and southern Oklahoma.

The estimated historical indigenous range (area of occupancy, AOO) of the whole species A. spinifera was 1,614,565 km², and the estimated historical indigenous extent of occurrence (EOO) was 6,550,365 km² (TTWG in press). The estimated historical indigenous range (AOO) of the subspecies A. s. spinifera was 1,017,323 km² (63.0% of the species' range), and the estimated historical indigenous range (EOO) was 4,377,847 km² (TTWG in press). For the subspecies A. s. aspera, the AOO was 250,732 km² (15.5% of the species' range), and the EOO was 696,348 km² (TTWG in press). For the subspecies A. s. atra, the historical AOO was 535 km² (only 0.03% of the species' range), and the EOO was 2,116 km² (TTWG in press). For the subspecies A. s. emoryi, the AOO was 145,362 km² (9.0% of the species' range), and the EOO was 645,870 km² (TTWG in press). For the subspecies A. s. guadalupensis, the AOO was 60,777 km² (3.8% of the species' range), and the EOO was 177,656 km² (TTWG in press). For the subspecies A. s. pallida, the AOO was 228,232 km² (14.1% of the species' range), and the EOO was 646,121 km² (TTWG in press). The current AOOs for all these subspecies and of the species as a whole are clearly less than the historical indigenous estimations noted here.

Apalone spinifera is a widespread and generally common species in the core of its range; however, some peripheral and marginal populations are sparse and appear to be declining. Depending on site and habitat type, the species constitutes between <1% to 67% numerically of all turtles encountered in turtle community surveys (reviews in Ernst et al. 1994, Ernst and Lovich 2009).

Detailed population monitoring data are scarce for A. spinifera, perhaps because of the perception that the species is resilient to habitat loss and modification. Rizkalla and Swihart (2006) found a positive correlation between A. spinifera occupancy and proximity to agricultural and urban areas. However, long-distance movements of A. spinifera have been attributed to habitat fragmentation (Galois et al. 2002), and significantly fewer A. s. emoryi were detected in two Texas counties in 2009 as compared to 1977, which may have been attributable to commercial harvest, urbanization at unprotected sites and land-use changes at protected sites (Grosmaire 1977, Brown et al. 2012, Mali et al. 2015). However, Plummer et al. (2008) found evidence of resilience in A. spinifera across 10 years in a small urban stream in Arkansas after a channelization event appeared to cause a moderate population decline (Plummer and Mills 2008). In contrast, a population of A. spinifera within a large urban centre in Ontario, Canada, over a 30-year study, was increasingly relegated to the shrinking natural areas remaining along the river, with oviposition, thermoregulation, foraging and nursery habitat consistently disturbed, damaged or lost, and multiple turtles found injured or killed due to fish hooks (S.D. Gillingwater, unpubl. data). The number of A. spinifera nests detected in 2015–2016 along a protected river in Louisiana was about half of those detected in 1993–1994, and may have been due to the heavy losses of nests in the latter period due to ATV traffic on nesting sandbars (Doody 1995, Godwin et al. 2021). The latter study prompted legislation designed to deter ATV traffic at the site. Population size and structure data (and analyses) are available for potential future monitoring in Arkansas (Plummer and Mills 2008), Montana (Tornabene et al. 2018, Ostovar et al. 2021), Pennsylvania (Mahoney and Lindeman, 2016), Texas (Brown et al. 2012), and Vermont (Parren et al. 2021).

Impoundments on the Missouri River system in North Dakota and South Dakota and Montana have led to local extirpations of A. spinifera and left populations in the upper Missouri River, Yellowstone River, and Bighorn River systems genetically and demographically isolated. In north-central Wyoming, the Bighorn River drainage has a peripheral population that was isolated from populations in Montana when the Yellowtail Dam was constructed in the 1960s. The Bighorn River population is currently of concern in Wyoming because the species is becoming rare. Although long-term monitoring data are not available, anecdotal evidence from the public and from biologists in the region suggests the population has declined. A recent concerted survey effort was conducted in 2021 by the Wyoming Game and Fish Department that resulted in extremely low catch-per-unit-effort (CPUE = 0.013 softshell turtles/trap night) across c. 140 km in the Bighorn River (Autz et al. 2024).

In Ontario, populations of A. s. spinifera have declined by up to 45% over the last 20 years due to a long list of threats, including habitat modification, water level management, ATV traffic, invasive species, livestock farming, illegal collection, boat collisions, and subsidized mesopredators (COSEWIC 2016, Ministry of the Environment, Conservation and Parks 2019). There are approximately 1,000 mature individuals; these currently occur in only 14 extant populations as compared to 35 historically recorded populations (COSEWIC 2016, Ministry of the Environment, Conservation and Parks 2019). In Quebec, A. s. spinifera is found only in Lake Champlain and its tributaries (Daigle et al. 2002, Galois et al. 2002). Historical or rare observations exist for two other rivers in southern Québec, from which they are considered extirpated (COSEWIC 2016). The disjunct Lake Champlain population of A. s. spinifera is shared with two U.S. states, Vermont and New York, and is separated by more than 200 km from the nearest population located in Lake Ontario. Studies in Canada span over 30 years, and may provide an indication of threats and impacts to the species over a larger extent of its range, though further long-term research in the USA and Mexico is necessary in order to better understand potentially similar threats across more of the species’ range. However, populations at northern range margins are impacted by a combination of a colder climate, which limits the species’ abundance, and anthropogenic threats, thereby possibly inflating those threats. In core distribution areas in warmer climates, A. spinifera has apparently been more resistant to anthropogenic threats such as habitat modification, fishing, and trapping.

In Texas, the status of the A. s. guadalupensis population has not been explicitly examined as a focal project species. It does, however, have a good amount of incidental data from work examining the turtle fauna and other reptiles of the region. Early characterizations of the species are found in works by Strecker (1922, 1926) and Strecker and Williams (1927). While those early authors did not report on any significant quantity of specimens, they did report the species as being considered abundant in local streams and rivers. Swanson (2009) surveyed historic wetlands of the San Marcos River and detected one individual among 119 turtles detected. M.R.J. Forstner and I. Mali (unpubl. data) have conducted turtle research in the Guadalupe River basin since 2000 and 2009–2016, respectively; they have captured, measured and released c. 3,500 turtles. Of that number, 396 were from within the distribution of A. s. guadalupensis within the Guadalupe River basin, and within that subset, only 26 (6.6%) were A. s. guadalupensis. Analogously, in recent surveys for Graptemys caglei by M.R.J. Forstner (unpubl. data, 2017–2024) within the Guadalupe River below Seguin, Texas, A. s. guadalupensis was observed at about 10% of the corresponding number of the generally scarce G. caglei, and comprised less than 1% of all the turtles detected. In the San Marcos River, A. s. guadalupensis made up fewer than 2% of all freshwater turtle detections across surveys from 2000–2010 (M.R.J. Forstner, unpubl. data; non-Apalone survey data from this work published in Bailey et al. 2014). Finally, in a recent turtle survey within the Guadalupe River system, Beverly (2023) reported 65 A. s. guadalupensis (1.3%) out of 5,192 total turtle detections. Thus, A. s. guadalupensis was historically reported as widespread and abundant, but currently available data do not support such abundance at present.

Also in Texas, the status of the A. s. emoryi population has been only minimally examined. In recent surveys of 17 sites in the lower Pecos River from the Red Bluff Reservoir to the river's confluence with the Rio Grande, Mahan (2022) did not detect the species at six localities. One obvious gap is located between the town of Grandfalls to the town of Iraan (c. 80 km straight line distance). In the Pecos River above Red Bluff Reservoir to Brantley Reservoir and at wetlands of Bitter Lake National Wildlife Refuge, New Mexico, A. s. emoryi is one of the most commonly encountered freshwater turtle species (Mahan 2022; I. Mali, unpubl. data). It is, however, exceptionally rare in the Black River (<1% of all turtle captures over eight years of intensive surveys), New Mexico tributary of the Pecos River, despite this river having higher overall quality of habitat. Apalone s. emoryi is one of the most common species of turtles in the Delaware River in New Mexico, also a tributary of the Pecos River, and is commonly encountered in hoop traps in the Rio Grande River (e.g., Big Bend National Park, Black Gap Wildlife Management Area) and wetlands of the Rio Grande Valley in southern Texas (Mali et al. 2013, 2015, but see Brown et al. 2011, 2012, for comparison with historical data). Population genetic work based on microsatellite loci indicates statistically significant differentiation between west Texas and south Texas populations (Mali et al. 2015). Dams on the Rio Grande (e.g., Amistad, Falcon) and Pecos River (Red Bluff, Brantley, Santa Rosa) likely represent a significant barrier to movement (Bárcenas-García et al. 2022) in addition to contributing negatively to overall water quality and habitat availability.

In Cuatro Ciénegas, Mexico, populations of A. s. atra are more abundant in deep wetlands such as the Mezquite River, but less abundant in shallow flooded areas such as lagoons and pozas (Cerdá-Ardura et al. 2008; G. Castañeda Gaytán pers. obs.).

In the southern portion of its range, Apalone spinifera occurs in most types of lotic and lentic wetlands, including rivers, streams, brooks, lakes, ox-bows, ponds, sloughs, bayous, ditches, canals, impoundments, borrow pits, desert springs, flooded gravel pits, swamps and freshwater marshes (reviewed in Webb 1962; Ernst et al. 1994; Anderson et al. 2002; Barko and Briggler 2006; Moler 2006; Tornabene et al. 2017, 2019). In the northern extent of the range, the species is less generalized in its habitat preferences, and is limited to rivers, appropriately-sized streams, lakes and rarely, glacial kettle ponds. This reduced habitat preference is likely due to dissolved oxygen requirements during long winter brumation periods in northern latitudes. Smaller lentic waterbodies directly adjacent to these larger systems may be used seasonally, outside of winter. A soft bottom with some aquatic vegetation appears required, as are sand bars or mudbanks for basking, and sand or sand and gravel bars directly adjacent to the waterbody for egg laying, while accumulations of underwater debris are preferred microhabitats. The species has been associated with highly modified environments; for example shorelines with reduced vegetation or channelization (Moll and Moll 2004, Plummer and Mills 2008, Hill and Vodopich 2013), though similarly modified environments have been shown to deter the species from use in Canada (S.D. Gillingwater, unpubl. data). The species also occurs in brackish water (Neill 1958, Webb 1962; J.S. Doody pers. obs.), and is reported from a salinity of 238 meq/litre (Seidel 1975), though such conditions are not common in the species’ range. When A. spinifera inhabits temporary wetlands it can readily move overland short distances when those habitats become dry, or move overland between wetlands (reviewed in Webb 1962, Williams and Christiansen 1981), though extensive overland movements have not been documented.

Adults preferred pools over riffles in a small stream in Arkansas (Plummer et al. 1997). Males and females can use different microhabitats in some systems (e.g., Williams and Christiansen 1981); however, they used the same microhabitats in a small creek in Arkansas (Plummer et al. 1997, Plummer and Mills 2008). The species shares habitats and microhabitats with A. mutica, but A. spinifera occupies more habitats (e.g., rivers, impoundments, temporary ponds) and microhabitats (i.e., with more submerged brush, fallen trees and other debris) than A. mutica, which is more often found in open water (Webb 1962, Doody 1996, Godwin et al. 2021, Williams and Christiansen 1981). In Ontario, Canada, it has been observed that adult females are less commonly located in shallow muddy bays and sandbars, compared with juveniles of both sexes and adult males, which extensively use these habitat types. Additionally, females have been found to make long-distance movements between habitat types (sometimes exceeding 35 km), whereas adult males made only small, local movements (S.D. Gillingwater, unpubl. data). Finally, the species shares basking microhabitats with A. mutica, but the latter are less likely to bask on logs, branches or other discrete objects (Williams and Christiansen 1981).

Apalone spinifera is omnivorous, but tends towards carnivory (Carr 1952, Webb 1962, Williams and Christiansen 1981). A dietary study of 52 stomachs from Iowa contained, in order of importance, by volume: crayfish, large fish, small fish, plant material, and aquatic insect larvae (Williams and Christiansen 1981); the plant material included acorns and leaves. Acorns and pecans have been found in the diet in Louisiana (Platt et al. 2008). Analysis of 29 stomachs from Minnesota contained insects and insect larvae, crayfish, sowbugs, snails, bivalves, earthworms, fish and vegetation (Cochran and McConville 1983). Analyses of smaller samples of turtles found, in general order of importance: crayfish, insects and fish (Newman 1906, Evermann and Clark 1920, Lagler 1943, Breckenridge 1944, Penn 1950, James 1966, Hopwood 1974, Minckley 1982). An analysis of faecal samples from Pennsylvania found that males ingested mainly insects and algal stalks while females consumed mainly algal stalks, crayfish and fish (Mahoney and Lindeman 2016). Lindeman (2000) predicted that the large-headed females of A. s. aspera in southeastern rivers would feed on bivalves, partly based on observations of captive females crushing the shells of Corbicula (J.S. Doody pers. obs.). Franklin (2014) reported an adult female feeding on a hatchling red-eared slider (Trachemys scripta) in Texas. Newly hatched neonates typically sample most items small enough to fit in their mouths, with sand and woody debris generally being spit out, but algae, plant matter and small invertebrates being consumed (S.D. Gillingwater, unpubl. data). Despite their generalist diet, turtles showed some evidence of selectivity compared to available prey in a study comparing stomach contents to dredge samples (Cochran and McConville 1983).

Carrion is taken; large fish in the stomachs probably were dead before consumption (Williams and Christiansen 1981); turtles in a fyke net fed upon dead fish (Cochran and McConville 1983), and an adult female was observed eating exposed flesh from the hind quarters of a dead white-tailed deer, Odocoileus virginianus, that was floating in a creek (Franklin et al. 2020). During an Ontario study along Lake Erie, A. spinifera, Chelydra serpentina, and Chrysemys picta marginata were all observed consuming a dead Freshwater Drum (Aplodinotus grunniens) at the same time (S.D. Gillingwater pers. obs.).

The occurrence of plant material in the diet, although typically less important in volume than invertebrates, appears to be deliberate (Mahoney and Lindeman 2016) rather than incidental (Cochran and McConville 1983). For example, stomach contents of three adults from an introduced population in Nevada revealed multiple fragments of Screwbean Mesquite fruits (Prosopis pubescens) and fruits and seeds of Desert Fan Palm, Washingtonia filifera (Heyborne and Sigg 2017). Surface (1908) reported corn as a usual food item in Ohio. J.S. Doody (unpubl. data) observed one adult male A. spinifera leave the water onto wet sand to retrieve Muscadine fruits (Vitis rotundifolia) that had fallen from an overhanging vine.  

Foraging can involve ambush predation with a gape and suck behaviour (Hudson 1942, Breckenridge 1944, Conant 1951) or active pursuit including thrusting the snout into vegetation or under rocks (Newman 1906). Larger prey can be held by the forefeet while torn into smaller pieces (Newman 1906). In Louisiana, J.S. Doody (pers. obs.) has regularly observed males foraging with male A. mutica in groups of 2–10, on the bottom in shallow, still-water areas at the ends of sandbars, especially in the evening. Compared to A. mutica, A. spinifera is primarily a bottom-feeder, based on its affinity for crayfish and carrion (Williams and Christiansen 1981). However, some terrestrial items (e.g., caterpillars, Muscadine fruits) found in some stomachs suggest surface and terrestrial feeding (Cochran and McConville 1983; J.S. Doody pers. obs.; see also Surface 1908). 

Males of A. spinifera mature at a relatively small size (8–10 cm plastron length (PL), about 11–14 cm carapace length (CL), 130 grams or more), while females mature at 18–20 cm PL, CL over 28 cm, and 15 kg (Webb 1962, Robinson and Murphy 1978). Maximum CL size of females is 54 cm CL, males maximum 27.5 cm. Maximum longevity is probably well over 30 years (Breckenridge 1955 in Ernst et al. 1994), and based on estimated age of maturity (12–13 years), longevity is estimated at about 60–70 years (Iverson 2024). Females normally produce two clutches per year, nesting mainly in June and July. Clutches usually comprise 12–18 eggs, but extremes of clutch size have been noted as 4–42 eggs (Ernst et al. 1994, Tamplin 2010) and clutches with as few as a single egg (S.D. Gillingwater, unpubl. data) and up to 43 eggs have been confirmed in Ontario, Canada (Gillingwater and Mackenzie 2015). According to growth models, males mature in their 4th or 5th year, whereas females mature in their 12th or 13th year, in Arkansas (Plummer and Mills 2015), and as such, generation time is estimated to be about 25 years (Iverson, 2024). 

In Louisiana, nesting is strongly bimodal, indicating that most females produced two clutches there (Doody 1995). At higher latitudes turtles may produce only one clutch, but some females from Quebec produced two clutches (Lazure et al. 2019), and in Ontario multiple females consistently produce two clutches a year, with second clutches notably smaller than the first (S.D. Gillingwater, unpubl. data). The estimation that the species produces 4–5 clutches annually in Iowa (Moll 1979, repeated in Iverson and Moler 1997) is probably in error, given that two clutches are produced annually in Louisiana (Doody 1995, Godwin et al. 2021). The narrow window of nesting in Montana suggests the production of one clutch annually there (Tornabene et al. 2018). 

Doody (1995) summarized the seasonal timing of nesting for A. spinifera across the range; most nesting occurred in May–July, with earliest record on 11 May and the latest on 28 July (see also Godwin et al. 2021). The nesting season may be more contracted at higher latitudes (e.g., Lazure et al. 2019), though in Ontario, Canada, oviposition has been confirmed as early as 26 May and as late as 23 July, depending on seasonal temperatures (S.D. Gillingwater, unpubl. data). The inclusion of August by Ernst et al. (1994) may be in error. Doody (1995) found no association between nesting date and body size (crawl width). The onset of nesting appears to be earlier in more southern (warmer) latitudes leading to a more prolonged nesting season (Doody 1995). In Quebec, nesting was more likely to occur when the difference between air and water temperatures was smaller (Lazure et al. 2019). Nesting typically ceases during flooding or discharge in riverine populations, possibly because all of the nesting grounds are underwater (Doody 1995, Tornabene et al. 2018, Lazure et al. 2019); similarly, in Ontario, nesting temporarily ceases during large flood events, during significant temperature drops, or during times of increased human presence near nesting areas, but generally resumes shortly after flood waters recede, temperatures rise, or humans leave the area. If flood waters remain high for multiple days, female turtles have been found to move to remaining higher elevation sites to deposit eggs, or nest in inappropriate areas where the eggs are likely to fail. In extreme cases, eggs have been found deposited in the water at the edge of flooded nest sites (S.D. Gillingwater, unpubl. data). The onset of nesting was delayed by discharge that flooded many island nesting areas in Montana (Tornabene et al. 2018). Flooding frequency changes due to climate change and dams can cause nest failure and the loss of sandbar nesting habitat (Dickerson et al. 1999). 

This species typically nests in open (canopy) areas free of shading vegetation, little or no ground cover, and that receive abundant solar radiation (Doody 1995, Dixon 2009, Dieter et al. 2014, Tornabene et al. 2018). Occasionally, nests are found in sparse vegetation (Doody 1995, Dixon 2009, Dieter et al. 2014, Tornabene et al. 2018). Nesting substrate is typically sandy, but nests can be found in gravel, gravelly sand, loamy sand, and other soils (Eigenman 1896, Newman 1906; Webb 1962; Doody 1995; Graham and Graham 1997; Tornabene et al. 2018, Parren et al. 2021). In Ontario, nests laid in clay or organic soils failed, with the eggs rotting, whereas those laid in sand or sand-gravel substrates completed incubation and young emerged. In multiple instances, plant roots invading the nest chamber also resulted in egg failure (S.D. Gillingwater, unpubl. data).

In a Louisiana river, crawl length for nesting averaged 57 m, including 35 m from the water to the nest (Doody 1995). Nests averaged 2.7 m above water in Louisiana (Doody 1995). Nests averaged 7 m from the Mississippi River in Minnesota (Pappas et al. 2017), 4–20 m from water in Winona, Minnesota (Vose 1964), 26 m from water in Louisiana (Doody 1995; includes A. mutica), and 61 m from water in Nebraska-South Dakota on natural sandbars (Dieter et al. 2014). In a small creek in Arkansas, four females nested within 3 m of the water (Plummer et al. 1997), whereas the species nested 1–15 m from water in Vermont (Graham and Graham 1997). However, on man-made sandbars, female Apalone spp. have been documented traveling up to 175 m from the water, likely due to man-made slopes having a lower average slope (2.8°; Dixon 2009). Nests close to the water on steep banks and sandbars revealed that height is the variable chosen by females and that distance is simply a function of height above water (Doody 1995). Nests were constructed in slopes averaging 17°, and tended to be in more south-facing aspects in Louisiana; however, this was an artefact of females choosing higher areas of sandbars which tended to be at the downstream ends of sandbars, rather than females choosing warmer aspects (Doody 1995). Over a 30-year study in Ontario, nests were located between approximately 1–20 m from water, along sand or gravel-sand banks, though at one site, a gravel road, over 20 m from the river, was used for nesting out of necessity. During flood events, the nests nearest the water would remain submerged longer, often leading to death of the eggs, whereas nests at higher elevations would often have higher survivorship if flood waters retreated within 24 hours. Alternately, nests at higher elevations often failed during drought conditions, with nests near water surviving. Survival was based on the given circumstances of the year, which varied considerably from year to year. Similarly, the choice of nesting location would sometimes be dictated by temperature and number of days between rain events, with turtles often nesting closer to water during drought conditions, and higher up the bank after a rain (S.D. Gillingwater, unpubl. data).  

Communal nesting is common in the species (Newman 1906; Evermann and Clark 1920; Minton 1972; Doody 1995; Graham and Graham 1997; Dieter et al. 2014; Godwin 2017; Tornabene et al. 2018, 2019; Parren et al. 2021). Traditional nesting, whereby females use the same nesting areas or nest sites across years, is probably common, but requires following individual nesting females across years to confirm. In Ontario, Canada, multiple communal nesting beaches have been confirmed, some with over 100 nests being laid on a single sand/gravel bar (S.D. Gillingwater, unpubl. data). Apalone spp. have a shorter incubation period than other turtle species (Janzen 1993) and hatching occurs from August through September. A South Dakota study found that on average there were 14 eggs in a clutch with an average depth of the top egg being 9 cm (Dixon 2009). 

Apalone spinifera atra is restricted to the Cuatro Ciénegas basin in Coahuila, Mexico, an hourglass-shaped intermontane basin of about 50 km long and 8–24 km wide (about 600 km²), its floor being at 720 m altitude (Cerdá-Ardura et al. 2008). Much of the central part of the basin is marshy, with dry sandy slopes leading up the rocky valley slopes. Several deep ponds (up to several metres) occur within the marshy area and retain crystal-clear water throughout the year. About half the bottom is covered by dense submerged aquatic vegetation (mainly Chara) and the other half is bare sediment. Waterlilies grow in the shallow parts, and thick stands of cattails (Typha) and Eleocharis fringe the ponds. The water temperature is about 27–29ºC. Ponds may be separated from dry nesting areas on the slopes by substantial distances (several 100 m) of flat marshy grassland (Webb and Legler 1960). Within the basin, A. s. atra has been recorded only within the deep ponds, and not in riverine situations (Webb and Legler 1960, Webb 1962, McGaugh and Janzen 2008, Cerdá-Ardura et al. 2008). Limited information of food indicates that the animals feed selectively on aquatic insect larvae (Webb and Legler 1960). Reproductive data on dissected type series (limited to smaller individuals) suggest that A. s. atra matures at similar sizes as other A. spinifera subspecies (Webb and Legler 1960), i.e., males mature at CL over 15 cm, and females at CL over 28 cm. No information is available on clutch size or frequency, on age at maturity, or longevity of A. s. atra. Field observations suggest nesting occurs in June and a dead hatchling with a CL of c. 40 mm was observed in August (G. Castañeda Gaytán pers. obs.).

Habitat loss, modification and fragmentation are collectively, likely the primary anthropogenic threats to Apalone spinifera. However, subsidized nest mesopredators (primarily raccoons) may be a more serious threat. Evaluating the impacts of these threats requires studies to determine the effects of predation rates on the population size and structure of populations. Turtle fraservirus 1 (TFV1) has not been detected or reported in A. spinifera, but has been implicated in A. ferox mortality events (Waltzek et al. 2023). Climate change effects, including rising nest temperatures and nest flooding, could decrease nest survival (Doody 1995, Godwin 2017, COSEWIC 2016). Water released from dams can cause nest failure and the loss of sandbar nesting habitat which could lead to lower recruitment and an overall population decline (Dickerson et al. 1999). In one study, all-terrain vehicles caused 34% nest mortality (Godwin et al. 2021, COSEWIC 2016). Like most aquatic species, A. spinifera can be affected by water pollution of waterways as it causes degradation of its aquatic habitat. Human harvest for both domestic and international markets occurs in the species (Ceballos and Fitzgerald 2004) and needs continued monitoring. At the individual and possibly population levels, drowning in nets (bycatch), fishhook ingestion, and boat strikes remain a problem (Barko et al. 2004, Galois and Ouellet, 2007, Fratto et al. 2008, Bury 2011, Steen et al. 2014, COSEWIC 2016, Lindeman 2018, Mahan et al. 2020, Ostovar et al. 2021).

Apalone spinifera has long been exploited for local consumption (Webb 1962) and more recently for the export of adults for food and of hatchlings as pets and for mainly Asian aquaculture operations (Mali et al. 2014). Some individuals are destroyed as nuisance by-catch by recreational fishermen or have long-term injuries such as fish hook ingestion (Steen et al. 2014), are run over when crossing roads, and populations are affected by pollution, water diversion, and water infrastructure development. The species as a whole does not appear threatened in its existence by present processes. However, certain subspecies and populations are considered to be less secure. For example, in Canada, A. s. spinfera is designated as Endangered both federally and provincially, owing to a range of threats that continue to impact populations across multiple river watersheds and lake ecosystems (COSEWIC 2016).

Apalone spinifera atra: The survival of this subspecies is dependent on the ecological and hydrological integrity of the relatively small Cuatro Ciénegas ecosystem. The Cuatro Ciénegas basin has been extensively altered in its hydrology by digging canals to supply water to a steel mill in Monclova, and for local agricultural irrigation. Roads, railroads, pipelines and other infrastructure for industrial, logistic and tourism and recreational purposes have impacted the ecosystem, and many of these environmental impacts continue. Another recorded threat is hybridization or intergradation with A. s. emoryi, which entered the Cuatro Ciénegas basin from the Rio Grande through the Rio Nadadores and Rio Salado. Several previously isolated ponds and wetlands had been drained and interconnected by a network of canals before 1960 (probably before 1920) for agricultural purposes (Webb and Legler 1960, Smith and Smith 1979), but apparently not connected to the Rio Chiquito. How A. s. emoryi transferred from the Rio Nadadores to the isolated ponds is unclear; Webb (1962) considered overland dispersal during rain, subterranean connections, and human transport as possibilities. Apalone spinifera emoryi was present in the Cuatro Ciénegas basin as early as 1938–1939 (Schmidt and Owens 1944); specimens of both forms collected in 1958 were clearly separable (Webb and Legler 1960) and showed no indication of hybridization at least 20 years after first contact. However, Webb (1962) subsequently noted that not all atra showed all diagnostic characteristics, and that some approached the condition shown in emoryi. Within the basin, emoryi shows a preference for riverine situations, with only few individuals recorded from ponds (Webb and Legler 1960, Webb 1962, McGaugh and Janzen 2008, Cerdá-Ardura et al. 2008). Some trade and consumption of softshell turtles has occurred in the Cuatro Ciénegas basin at least in the 1950s (Webb 1962).

Apalone spinifera emoryi: Populations in Mexico are impacted by water diversion and reduced groundwater levels from irrigation and groundwater pumping. Populations in Texas and New Mexico have been impacted to varying degrees by urbanization, river flow alteration, and direct harvest (Ceballos and Fitzgerald 2004, Brown et al. 2012, Mali et al. 2015). Dam construction and the oil and gas industry have contributed to diminished flows and increased salinization of the lower Pecos River below Red Bluff Reservoir. Recent surveys of 17 sites in the lower Pecos River from the Red Bluff Reservoir to the river's confluence with the Rio Grande (Mahan 2022) did not detect species at six localities, with an obvious gap between Grandfalls and Iraan (c. 80 km straight line distance). These areas have conductivity of over 20,000 µS/cm, exemplifying that some taxa cannot tolerate extremely saline segments of the Pecos River. In addition, several of the 17 sites sampled had obvious crude oil residues, pointing to pollution issues. There is a lack of systematic surveys that are explicitly targeted to study population trends other than more than a decade old studies in the Rio Grande Valley of South Texas by Brown et al. (2011, 2012) that examined the effects of unregulated harvest and its extreme pressures on wild A. s. emoryi in this region (Ceballos and Fitzgerald 2004). Based on this work, it is clear that unregulated harvest has had a significant negative impact on the taxon.

Apalone spinifera guadalupensis: The threats to this central Texas endemic subspecies are similar to those affecting A. s. emoryi in southern Texas.

Large numbers of adults (mainly females) were exported from the USA to East Asia from the late 1990s until the 2010s, as well as large quantities of hatchlings (from farms/ranches and wild-harvested eggs); this trade has fluctuated as East Asian jurisdictions reduced imports in response to veterinary concerns and the maturation of domestic turtle aquaculture operations. 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 Clemmys serpentina and/or Apalone spp.).

Declared exports of Apalone spinifera are recorded in the USFWS Law Enforcement Management Information System (LEMIS) database, but available numbers do not reliably allow separating the data by subspecies, area of origin, or whether collected from the wild or produced in captive conditions. Numbers of live specimens exported from the USA were under 10,000 per year during 1999–2002, rising steadily to a peak of nearly 130,000 in 2007, after which trade volumes declined to between 47,006 and 75,728 for the years 2009–2013, and further declined to under 1000 animals per year from 2016 onwards. In total, 715,499 live specimens were declared to have been exported during 2000–2019 (see attached Supplementary Information for annual declared numbers of A. spinifera, A. ferox, and A. mutica), in addition to some quantities of meat and eggs that were also recorded as exported.

Apalone spinifera is managed as a non-game resource in much of the USA and occurs in a wide variety of sites and habitats under various degrees of protective measures (Nanjappa and Conrad, 2011). The species was listed in CITES Appendix II as Apalone spp. in 2023, which should allow for better population monitoring and understanding of harvest levels and population dynamics. The species was previously listed in CITES Appendix III (USA) from 2016 to 2022. In Minnesota, A. spinifera was banned from harvest in 2022 partly due to concern about potential overharvest, but also due to the likelihood that A. mutica was being illegally taken due to improper identification of the species (C.D. Hall pers. comm.). In Vermont, legislation has been passed to designate and protect “critical habitat” for threatened and endangered species, and four nesting beaches for A. spinifera have been designated and protected (L. Groff pers. comm.). Texas has banned commercial harvest of the species, effective 2018.

According to NatureServe (2024), this species is ranked as Secure (G5) because of its large range with many occurrences and no evidence of drastic declines. However, Minnesota, South Dakota, Georgia, Florida, and North Carolina have it ranked as Vulnerable (S3); Virginia, New York, and Ontario consider it Imperiled (S2), whereas it is listed as Critically Imperiled (S1) in Maryland, Vermont, and Quebec. Based on the data available to them, NatureServe (2024) has suggested a decline of <30% to an increase of 25% in long-term population trends of this species. 

The subspecies Apalone spinifera atra has been included in Appendix I of CITES since 1975, prohibiting any form of commercial international trade, and is protected from exploitation under Mexican wildlife and natural resource legislation. Its entire range falls within the 843 sq. km Cuatro Ciénegas Flora and Fauna Protection Area (IUCN Category VI), established in 1994, and its current AOO is estimated at significantly less than 500 km². Much more information on the species’ natural history is urgently needed, and an investigation of the ecological and genetic effects of the presence of invasive A. spinifera emoryi in Cuatro Ciénegas is a top priority, with work by McGaugh and Janzen (2008) demonstrating hybridization between the two taxa.

In Canada, A. spinifera is listed as Endangered under the federal Species at Risk Act, as Endangered under the Ontario Species at Risk Act (2007), and Threatened in Quebec under the Act respecting threatened and vulnerable species (R.S.Q., c. E-12.01). Additionally, it is designated as a specially protected reptile under the Ontario Fish and Wildlife Conservation Act. Protections from the various legislations in Canada make it illegal to damage habitat, or to take, harm, harass, keep, or sell the species. Though threats persist, such sweeping protections reduce overall impacts on the species, and provide legal options when necessary. The disjunct populations of A. s. spinifera in Lake Champlain and the Hudson River Valley are considered to warrant focused conservation attention, with the populations in Vermont rigorously protected.