Thomas et al. (2014) described the Suwannee alligator snapping turtle (Macrochelys suwanniensis) as distinct from the western alligator snapping turtle (Macrochelys temminckii) based on genetic differentiation and differences in skull and carapace morphology. Description of this new species was supported by a subsequent analysis of cranial shape (Murray et al. 2014), a review of population phylogenetic structure of Macrochelys, and an evaluation of morphological and molecular data (Folt and Guyer 2015). Macrochelys suwanniensis is most readily distinguished from other Macrochelys by a wide and lunate carapacial caudal notch that usually includes the pygal and peripheral pair 11 (Thomas et al. 2014).

The conservation of Macrochelys as a whole is of significant concern (Pritchard 1989; Reed et al. 2002) and the US Fish and Wildlife Service (USFWS) was petitioned in 1983 and in 2012 to federally list M. temminckii. The USFWS proposed that M. suwanniensis be listed as Threatened under the Endangered Species Act (USFWS 2021). Florida prohibited sale of Macrochelys in 1972 and limited personal possession to 1 turtle in 1973, effectively ending commercial harvest. Georgia prohibited commercial harvest in 1989 and listed Macrochelys as State Threatened in 1992. Florida prohibited all take of Macrochelys in 2009 and listed M. suwanniensis as State Threatened in November 2018. In a species status assessment for M. suwanniensis, the current population size is estimated to be 2000 adults, with 50% confidence that the true size is between 500 and 5000 adults (USFWS 2020).

The distribution of M. suwanniensis is apparently restricted to the Suwannee drainage in southern Georgia and northern peninsular Florida, which encompasses an estimated 26,641 km² (Nordlie 1990). As recently as 100,000 yrs ago, Macrochelys occurred farther south in the Florida peninsula and at sites that currently drain into the St. Johns River (Atlantic drainage), but saltwater inundation during elevated sea levels reduced its range (Ewert et al. 2006). Jensen and Birkhead (2003) unsuccessfully trapped for Macrochelys in the St. Mary's River along the Georgia/Florida state line in the Atlantic drainage. The distribution and relative abundance of M. suwanniensis in portions of the Suwannee drainage are unknown. Based on observations and limited trapping, Pritchard (1989) claimed Macrochelys was rare in the 200,000-ha Okefenokee Swamp and upper Suwannee River, and he speculated that populations may have declined sharply in the Suwannee River since the 1930s due to overharvest. Trapping pressure on Macrochelys was intense until approximately 1973 in the upper Suwannee River as far upstream as the entrance to Stephen C. Foster State Park (Pritchard 1989) and populations may not have recovered from past harvest.

The Suwannee River originates as a blackwater stream in the Okefenokee Basin of southeastern Georgia and is fed by blackwater streams, spring-run streams, and springs as it flows approximately 378 km southwesterly toward the Gulf of Mexico in Florida. The Suwannee River has a mean flow of 305 m³/s (Nordlie 1990) and experiences changes in water chemistry and biological productivity as it flows through Florida (Ceryak et al. 1983), leading the Suwannee River Water Management District to divide the main stem of the river into 6 distinct ecological reaches: Upper River Blackwater (Reach 1), Cody Scarp Transitional (Reach 2), Middle River Calcareous (Reach 3), Lower River Calcareous (Reach 4), Tidal Riverine (Reach 5), and Estuarine (Reach 6; Hornsby et al. 2000). The Santa Fe River is the major Florida tributary of the Suwannee River, and its tributaries include the New River, Olustee Creek, and the Ichetucknee River. The first 2 tributaries are blackwater streams, whereas the Ichetucknee River is a spring-run stream only 9.7 km long. The Santa Fe River begins as a blackwater stream in swamps near Lake Santa Fe and Lake Alto and becomes increasingly spring-fed as it flows 115 km to the Suwannee River. The Santa Fe River disappears underground for about 5 km in River Rise Preserve State Park, approximately 60 km from its origin. Johnston et al. (2015) extensively trapped the Santa Fe River upstream (upper) and downstream (lower) of this land bridge and described how the river changes from a blackwater stream after it receives input from at least 45 artesian springs (Scott et al. 2004), increasing its water clarity, thermal stability, and mineral content. Moler (1996) reported earlier trapping efforts in the Santa Fe drainage.

Jensen and Birkhead (2003) trapped in the Suwannee drainage in Georgia, finding Macrochelys in the Alapaha, Little, and Withlacoochee rivers but not in the main stem of the Suwannee River or in Suwannoochee Creek. The Alapaha and Withlacoochee rivers are blackwater with narrow floodplains and sections of these streams flowing through karst terrain disappear underground into sinkholes during periods of low water (Wharton 1978; McConnell and Hacke 1993; Torak et al. 2010; Edwards et al. 2013).

To remedy the lack of knowledge regarding the distribution and relative abundance of M. suwanniensis and because of its imperiled status, we compiled existing records and trapped for Macrochelys throughout the entire Suwannee drainage plus streams in the Big Bend region of Florida south and north of the Suwannee River. The lack of Macrochelys records from streams between the Suwannee and Ochlockonee rivers apparently represents a genuine distributional gap in the range (Allen and Neill 1950), but a few reports exist from some of these streams (Pritchard 1989; Ewert et al. 2006), We used the number of turtles captured per trap night as an index of relative abundance and speculate on reasons for observed differences in relative abundance among streams and sections of streams.

METHODS

We compiled M. suwanniensis museum records, literature, and personal communications, as well as catch per unit effort (CPUE) data from a review of this literature, including unpublished reports and our trapping surveys. Museum records were obtained from the American Museum of Natural History (AMNH), Auburn University Museum, Georgia Museum of Natural History (GMNH), Georgia Southern University (GSU), and Florida Museum of Natural History, University of Florida (UF).

From July 2011 through August 2020, we trapped for M. suwanniensis using 122-cm-diameter hoop net traps with 6.4-cm mesh. However, 76-cm-diameter hoop net traps with 2.5-cm mesh were used in the Ichetucknee River in Florida because smaller turtle species were the target. Georgia traps were baited with canned sardines and/or chunks of catfish, mullet, and tilapia, while Florida traps were baited with thawed chunks, heads, or fillets of various freshwater or marine fish species. We set traps during the afternoon or evening in water 0.5–2.5 m deep, typically upstream of logjams, large submerged logs, submerged trunks or roots of bald cypress (Taxodium distichum), steep vertical or undercut banks, or deep pools associated with outer bends in the river. We sometimes were forced to set traps based solely on suitable water depths, regardless of whether suitable downstream refugia were apparent. Traps were set parallel to the bank with the funnel opening facing downstream and a portion of the upstream end of the trap above water, leaving an air space. In estuarine sites, we usually set paired traps (counted as 1 trap) facing upstream and downstream so that they could effectively trap when the direction of the current reversed due to tides. In Florida, we checked traps the following morning and then moved them to a different section of the river. In Georgia, we checked traps daily but left them in place for up to 3 nights. The number of turtles captured per trap night (TN) is referred to as CPUE.

Downstream of White Springs, Florida, we selected two 5-km stretches in each of 6 ecological reaches of the Suwannee River and trapped them multiple times using 12 traps per session (i.e., 1 night of trapping) for a total of 766 TN. Elsewhere, we did not standardize the number of traps set, number of trapping sessions, or length of stream trapped. Site selection was usually determined by accessibility and some small streams were trapped only once with a few traps. In the Santa Fe drainage, we compiled 504 TN by combining our data with those of Johnston et al. (2015) and eliminating results from smaller traps, except in the Ichetucknee River (Table 1). The only other streams with more than 100 TN were the Alapaha and Suwannee rivers in Georgia (Table 1). We did not trap the Okefenokee Swamp because of the presence of numerous large American alligators (Alligator mississippiensis).

Table 1. Trapping results from the Suwannee drainage in Georgia and Florida. TN = number of trap nights; CPUE = number of turtles captured per trap night.

View Fullscreen

For each turtle capture, we measured median straight-line carapace length (SCL) with Haglof aluminum tree calipers and precloacal tail length (PTL). Turtles were weighed using a Rubbermaid Pelouze 100-kg hanging scale or a Pesola 10-kg hanging scale. These measurements were used to determine age class and sex (Dobie 1971). We marked each turtle individually by drilling holes in marginal scutes (Cagle 1939) and/or by implanting a BioMark passive integrated transponder (PIT) tag into the ventral tail muscle. Most turtles were photographed, and significant records were photo vouchered in herpetology museum collections at GMNH, GSU, and UF. No specimens were collected, but tissue samples were acquired for genetic analyses. We obtained necessary permits and followed guidelines from the Herpetological Animal Care and Use Committee of the American Society of Ichthyologists and Herpetologists in handling and processing turtles.

RESULTS

Distribution. — We compiled 111 museum records, 16 literature records, and 40 records based on credible observations or photographs of M. suwanniensis (Appendix 1). Records with imprecise locality data were not mapped for Georgia (Fig. 1) and Florida (Fig. 2), unless they represented the only records for the waterbody. We did not include in Appendix 1 or map every turtle trapped but included at least 1 record from every 5-km river segment trapped. Florida accounted for 110 records and Georgia for 57 records (Appendix 1). Records spanned the period 1912–2020, with 124 records (74.2%) from the past 20 yrs (Appendix 1; Figs. 1 and 2). We trapped 267 turtles in Florida and 54 turtles in Georgia.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

We located a M. suwanniensis shell (UF 190009) on display at a fish camp on the west side of the Okefenokee Swamp near Fargo, Georgia, that came from a dead adult found in the 1980s by A. Griffis. In December 2019, a large male M. suwanniensis was photographed (GMNH 52076) in the Suwannee Canal near the Richard S. Bolt Visitor Center (site of the former Camp Cornelia) in the eastern section of the Okefenokee Swamp National Wildlife Refuge near the confluence of the Suwannee Canal and Cornhouse Creek, a tributary of the St. Mary's River. These findings indicate the species is still present in the Okefenokee Swamp, which we did not trap. We failed to capture M. suwanniensis in the Suwannee River in Georgia (Fig. 3) and at 2 sites in Florida upstream of White Springs (Fig. 4). However, B. Johnson photographed an adult M. suwanniensis (UF 190014) that he snagged in 2017 while fishing in the Suwannee River at the mouth of Tom's Creek 200 m south of the Georgia line (Fig. 1), which represents our only record from the main stem upstream of White Springs, Florida (Appendix 1).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Macrochelys is found in several tributaries between the Okefenokee Swamp and White Springs, Florida (Table 1; Appendix 1). We trapped 2 adult males in Suwannoochee Creek (10 TN). In 2016, C. Davis photographed an adult M. suwanniensis (UF 190012) that was probably nesting on a primitive road adjacent to Jones Creek in Clinch County, Georgia (Fig. 1). We had 5 captures of 4 unique individuals in 11 TN in Tom's Creek in Georgia just north of the Florida line (Fig. 3). We trapped 2 adult females in 6 TN in Rocky Creek (Fig. 4), a small tributary about 6 km south of the Georgia line in Hamilton County. In October 2016, J. Willmott observed 3 large M. suwanniensis (UF 189493) together in a shallow pool in Hunter Creek. When we trapped Hunter Creek (3 TN), the water was too shallow for effective trapping or to provide good cover for turtles.

In Georgia, we trapped M. suwanniensis in the Alapaha and Withlacoochee rivers and several of their tributaries: the Alapahoochee, New and Willacoochee rivers and Okapilco, Piscola, and Warrior creeks (Table 1; Fig. 3). We vouchered the first record from the Alapaha River in Florida (UF 191014), which was the smallest turtle trapped during our study (105 mm SCL). The farthest upstream record is from Warrior Creek, a tributary of the Little River, near Sylvester, Worth County, Georgia (Fig. 1).

In Florida, we trapped M. suwanniensis in all 6 ecological reaches of the Suwannee River, the upper and lower Santa Fe River, and the following tributaries of the Santa Fe River: Olustee Creek and the New and Ichetucknee rivers (Table 1; Fig. 4). The first voucher from Olustee Creek (UF 190977) was a large male (597 mm SCL; 45 kg) that had been trapped 10 yrs earlier 1 km downstream at its confluence with the Santa Fe River. We also vouchered the first specimen (UF 191028) from the New River, but unvouchered records already existed from there (Appendix 1). In 1982, an angler caught a Macrochelys in Alligator Lake (about 320 ha) in Lake City, Columbia County (UF 65902), which is closest to the headwaters of Price Creek, a tributary of Olustee Creek.

Relative Abundance. — Based on trapping success, relative abundance is higher in the middle Suwannee River (Reaches 2–4; CPUE = 0.24–0.38; 391 TN) than in the upper (Reach 1; CPUE = 0.19; 124 TN) and lower (Reach 5; CPUE = 0.18; 131 TN) main stem of the river (Table 1; Fig. 4). Relative abundance is very low in the estuary (CPUE = 0.01; 120 TN) and upstream of White Springs in Florida and Georgia (CPUE = 0; 174 TN) (Table 1). However, many variables could have affected trapping success in the estuary, including the wide width of the river and variable tidal flow. Relative abundance is similar in the Suwannee River downstream of White Springs (excluding the estuary; mean CPUE = 0.25; 646 TN) and in the upper Santa Fe River (CPUE = 0.26; 200 TN), but abundance is apparently lower in the lower Santa Fe River (CPUE = 0.09–0.12; 309 TN; Table 1). Overall, relative abundance in tributary streams is apparently lower in Georgia than in Florida. We had sufficient trapping effort in Georgia to conclude that relative abundance is higher in the Alapaha River (CPUE = 0.15; 199 TN) than in the Withlacoochee River (CPUE = 0.06; 67 TN). We could not conclude that some streams lacked M. suwanniensis or had a low or high relative abundance because trapping effort was insufficient or conditions were not conducive for successful trapping because of water levels, stream current, poor cover, or other factors.

DISCUSSION

Distribution. — The status of M. suwanniensis in the Okefenokee Swamp is poorly known and enigmatic. Various authors (e.g., Wright and Funkhouser 1915; Laerm et al. 1980) listed the alligator snapping turtle as an inhabitant of the Okefenokee Swamp. Pritchard (1989) stated that Macrochelys was rare in the Okefenokee Swamp and the upper Suwannee River. Following extensive wildfires in the mid-1950s, the headwaters of the Suwannee River were impounded by an earthen dam, the Suwannee River Sill, in 1960–1962, causing some changes to swamp hydrology and vegetation. The maximum water level of the swamp at the sill during zero outflow conditions has risen 12 cm and extends 60 km into the swamp (Malcolm et al. 1994). The palustrine environment of the sill impoundment differs from the naturally dynamic hydrology of a riparian environment (Yin and Brook 1992; Loftin et al. 2000). Construction of this sill may have caused changes to fish species abundance and composition in the upper Suwannee River (Wharton 1978), but its effects on M. suwanniensis populations and movements are unknown. The only museum specimen from the Okefenokee Swamp was a skull collected in 1912 by one of the Cornell University expeditions (AMNH 69731). Carr (1952) provided 2 photographs of a female from the Okefenokee Swamp, which we did not include in Appendix 1 because of imprecise locality information. A photograph of a large male in 2019 represents only the second voucher specimen from the swamp (Appendix 1), although De Sola and Abrams (1933) mentioned that alligator snapping turtles were “frequently hooked in the old drainage canal (St. Mary-Suwanee Canal) near Camp Cornelia in the Okefinokee.” Macrochelys suwanniensis apparently does not occur in rivers in the St. Mary's drainage (Allen and Neill 1950; Pritchard 1989; Jensen and Birkhead 2003), but the eastern portion of the Okefenokee Swamp, where Camp Cornelia is located, is in the St. Mary's drainage (Edwards et al. 2013). No streams cross Trail Ridge, a Pleistocene shoreline that forms the eastern boundary of both Okefenokee Swamp and the eastern geographic limit of M. suwanniensis (Pritchard 1989).

Like Jensen and Birkhead (2003), we failed to trap M. suwanniensis in the Suwannee River upstream of White Springs. However, we documented turtles in 5 tributaries in this section of the river: Jones, Suwannoochee, Tom's, Rocky, and Hunter creeks. We trapped M. suwanniensis in Suwannoochee Creek, where Jensen and Birkhead (2003) were unsuccessful but A. Redmond, a commercial trapper, reported the species was “fairly common” (Pritchard 1989). These tributaries are fringed by mixed pine–oak forest communities and offer cooler water temperatures than the river main stem. The narrower width of tributaries may help turtles locate traps, increasing the likelihood of our catching them.

Macrochelys populations are present in the Alapaha drainage in Georgia and Florida, including its tributaries, the Alapahoochee and Willacoochee rivers (Table 1; Fig. 1). Dozens of deep, natural swamp lakes pockmark the channel of the Alapaha River and would appear to offer prime habitat for M. suwanniensis, which has been documented from swamp lakes in the Alapaha River Wildlife Management Area. In Georgia, the Withlacoochee River and a large tributary, the Little River, have M. suwanniensis (Pritchard 1989; Jensen and Birkhead 2003). We failed to trap turtles in the Little River (Table 1; Fig. 3) but located observations and 1 museum voucher (GMNH 52117) from 2 sites in the past 5 yrs. Based on conversations with A. Redmond, Pritchard (1989) reported that intense commercial trapping occurred in the late 1960s and early 1970s in the Little River and that the Withlacoochee River was “trapped out” by 1982. Populations have at least partially recovered from past harvest in these 2 rivers (Table 1). We trapped M. suwanniensis in 2 smaller Georgia tributaries of the Withlacoochee River, Okapilco and Piscola creeks (Fig. 3), where Jensen and Birkhead (2003) were unsuccessful (Table 1). Three M. suwanniensis were caught by D. Winterton in the Withlacoochee River in northeastern Madison County, Florida (Pritchard 1989), and we trapped 2 M. suwanniensis in this river just north of the Georgia state line (Table 1; Fig. 3).

Johnston et al. (2015) trapped extensively in the Santa Fe River downstream of its confluence with the New River, but the distribution of M. suwanniensis farther upstream is unknown. We trapped the species in its 2 major tributaries, the New River and Olustee Creek, near their confluence with the Santa Fe River (Fig. 2). Turtles have been reported (Pritchard 1989), trapped (Johnston et al. 2015), and observed nesting along the Ichetucknee River, and nests have been found along Cow Creek, a tributary of the Santa Fe River in Gilchrist County (Jackson and Thomas 2018). The species is occasionally also found in springs and sinkhole lakes in the Santa Fe drainage (Johnston et al. 2015).

Voucher specimens are lacking from the 7 streams between the Suwannee and Ochlockonee drainages (Fig. 2), but reports exist from this purported distributional gap. Five of these streams begin as blackwater streams, but the Wakulla and Wacissa rivers originate at first-magnitude springs. The mouths of the Steinhatchee and Suwannee rivers are separated by about 50 km, whereas the mouths of the Ochlockonee River, which is inhabited by a different Macrochelys species, and the St. Marks River are separated by about 14 km. Pritchard (1989) reported 2 Macrochelys sightings from the Wacissa River and 1 sighting from Wakulla Springs. Florida Natural Areas Inventory records include a photograph in the Woods and Water newspaper of a specimen reportedly from the 111-km-long Aucilla River in 2000, a large dead adult from the St. Marks River in 1992, and a live specimen observed in the Wakulla River in 1980. Moler (1996) provides anecdotal reports of 1 additional specimen each from the St. Marks and Wakulla rivers but failed to trap Macrochelys in the Aucilla and St. Marks rivers. We did not map any of these observations because the lack of voucher specimens prevents identification to species. We conclude that Macrochelys populations are probably not established in these streams, despite the presence of suitable habitat. Some of these reports are credible, and turtles may occasionally wander in from the Ochlockonee River, the nearest river with an established Macrochelys population. Macrochelys has been observed in brackish Ochlockonee Bay (Pritchard 1989). Alternatively, turtles found in these streams could represent released pets. We are aware of Macrochelys that have been found outside their native range in Georgia (Fulton County) and in Florida: DeSoto, Duval, Hillsborough, Lee (UF 191797), Marion (AMNH 125696, KU 61844), Miami-Dade, Nassau, Orange (UF 166530), and Volusia counties.

Macrochelys suwanniensis is currently distributed widely throughout southcentral Georgia in medium- to large-sized streams (Fig. 1), with our surveys documenting a number of new county records and minor range extensions (Stevenson et al. 2021). We found evidence that the species occurs in every major creek in the Suwannee drainage in Florida. We documented the first records from the New and Willacoochee rivers and Okapilco, Piscola, Warrior, Jones, and Toms creeks in Georgia and from Rocky and Olustee creeks in Florida.

Relative Abundance. — Macrochelys suwanniensis abundance is apparently low in the Okefenokee Swamp and the upper Suwannee River north of White Springs, Florida (Pritchard 1989; Jensen and Birkhead 2003), and we failed to find the species during 130 TN in this section of the river but trapped it in several small tributaries. Macrochelys may have always been scarce in the upper Suwannee River (Jensen and Birkhead 2003) or populations may not have recovered from past harvest. Its low abundance in the main stem may be due to low pH and scarcity of potential prey (Jensen and Birkhead 2003). Mussels and snails, important food items in the diet of M. temminckii in Louisiana drainages (Elsey 2006), are absent from the upper Suwannee River (Williams et al. 2014).

In the Suwannee River, relative abundance is highest in the middle reaches, where increasing amounts of water from the Floridan Aquifer change the acidic, blackwater stream to a clear, slightly colored, alkaline stream with higher biological productivity (Hornsby et al. 2000). The paucity of Macrochelys in the estuary (Table 1) may be due to reduced available habitat or avoidance of water with higher salinities; the only capture was near the mouth of a small tributary (Fig. 4) that presumably had lower salinity. Adult Macrochelys are sometimes found in brackish-water habitats (Jackson and Ross 1971), but movements into salt water are extremely rare (Pritchard 1989; Ewert et al. 2006).

Based on CPUE, M. suwanniensis appears to be relatively abundant in some of the small tributaries off the upper Suwannee River, but this might be an artifact of turtles being more trappable in narrower streams. Another factor affecting this abundance metric is the number of trap nights; the capture of just 1 turtle can result in a high CPUE if trapping effort is low. Excluding trap nights in the estuary, our CPUE downstream of White Springs in the Suwannee River was 0.25, which we consider to be a baseline for medium- and large-sized streams such as the Alapaha, Withlacoochee, and Santa Fe rivers. In comparison, recent trapping for M. temminckii in 29 streams of all sizes in the Florida panhandle between the Sopchoppy and Perdido rivers, excluding the Choctawhatchee drainage (documented to have depauperate Macrochelys populations), yielded a CPUE of 0.30 in 586 TN (Enge et al. 2019). Multiple trapping sessions in three 5-km sections of the Ochlockonee, Apalachicola, and Choctawhatchee rivers in Florida yielded CPUEs of 0.53 (98 TN), 0.36 (107 N), and 0.01 (103 TN), respectively (Mays et al. 2015). Trapping for M. temminckii in Georgia by Jensen and Birkhead (2003) yielded a mean CPUE of 0.20, but the CPUE was 0.45 in several streams in the Apalachicola drainage. Our CPUE in several Georgia streams would probably have been higher if we had not left traps for more than 1 night, because most turtles were trapped the first night.

Relative abundance of M. suwanniensis in Georgia is relatively low. Sufficient trapping has been conducted in the Alapaha and Withlacoochee rivers to provide meaningful results. Interestingly, we had higher CPUE than Jensen and Birkhead (2003) in the Alapaha River, but this trend was reversed in the Withlacoochee River (Table 1). Heavy harvest apparently did not occur in most streams inhabited by M. suwanniensis, unlike many rivers in the Southeast, but harvest occurred in Georgia in the Little, Withlacoochee, and upper Suwannee rivers and Suwannoochee Creek (Pritchard 1989). These streams have low abundance of M. suwanniensis (Table 1), possibly indicating slow population recovery from past harvest. Because of long generation times and low reproductive rates of Macrochelys (Tucker and Sloan 1997; Reed et al. 2002; Holcomb and Carr 2013), depleted populations may take a long time to recover (King et al. 2016). Exceptionally large turtles (45–57 kg; old adult males) are present in the Suwannee and Santa Fe rivers in Florida (Enge et al. 2014; Johnston et al. 2015), which were probably never intensively harvested commercially. Although we did not analyze demographic data, we trapped juveniles (105–325 mm SCL) in Okapilco, Olustee, Tom's, and Warrior creeks and in the Alapaha, Santa Fe, Suwannee, and Withlacoochee rivers, indicating population recruitment is occurring in both states and in all drainages. We had recent reports of juveniles from the Alapahoochee and Little rivers in Georgia. Because Macrochelys does not attain sexual maturity until 11–22 yrs old (Dobie 1971; Tucker and Sloan 1997; Johnston et al. 2012), juveniles should be detected in those streams with substantial trapping effort, although different micro-habitat use and foraging strategies may make juveniles less susceptible to trapping than adults.

Macrochelys is relatively abundant in the Santa Fe River (Johnston et al. 2015) and its major blackwater tributaries, the New River and Olustee Creek (Table 1). Based on CPUE, the Suwannee River downstream of White Springs and the upper Santa Fe have similar abundance, but abundance is apparently lower in the lower Santa Fe River (Table 1). The species is less abundant in the short, spring-run Ichetucknee River (Johnston et al. 2015). During a 4-yr study in the upper Santa Fe River and New River, S. Santhuff claimed that he consistently had a CPUE of 0.30–0.60 (Moler 1996). Johnston et al. (2015) had a much lower CPUE in the upper Santa Fe River, but it was over twice as high as in the lower Santa Fe River (Table 1). We had a CPUE of 0.50 in the New River (Table 1), but this result should be interpreted cautiously because we had only 6 TN.

Threats. — Although harvest of M. suwanniensis is now illegal, anthropogenic mortality continues from harvest for food or pets, accidental ingestion of fishing tackle, and boat strikes (Ewert et al. 2006; USFWS 2020). Thirty-six percent of 11 radiographed Macrochelys in the Santa Fe River and 12% of 25 turtles in the Suwannee River in Florida had hooks lodged in their upper gastrointestinal tracts (Enge et al. 2014). Turtles are susceptible to getting caught by trotlines and bush hooks (i.e., limb lines) set for catfish or by recreational anglers (Enge et al. 2014). The monofilament or gel-spun fishing line attached to a hook can cause severe digestive blockage, and ingested hooks can perforate the digestive tract lining, resulting in injury and potentially death. Macrochelys individuals occasionally drown from entanglement in fishing line, particularly unattended bush hooks or trotlines (Mays et al. 2015). We had 3 reports of anglers in Georgia catching M. suwanniensis while fishing during the day, and 2 marked turtles in the Suwannee River in Florida were caught by anglers fishing for catfish at night from the riverbank. Based on published life-history data, mortality of adult Macrochelys from ingested hooks is sufficient to cause population declines (Steen and Robinson 2017). Macrochelys populations are susceptible to declines from anthropogenic mortality and are slow to recover because of low recruitment, slow growth, and long generation time (Iverson 1991; Congdon et al. 1994; Folt et al. 2016).

Groundwater withdrawal and water pollution pose potential threats to M. suwanniensis. Water levels in the Floridan Aquifer have declined in northern Florida over the past 70 yrs because of groundwater extraction for human use and reduced groundwater recharge due to surface drainage alterations (Knight 2015). Most of the 197 springs identified in the Suwannee River basin are located between Suwannee Springs and Ellaville in Florida. Reduced input from springs can decrease water levels and clarity and reduce biological productivity, potentially affecting turtle populations. In the Santa Fe River, the total density of all turtle species combined appeared to be 3 to 4 times greater in the spring-influenced reach than in the blackwater reach (Johnston et al. 2016). The area around the Suwannee River is sparsely populated, and the predominant land uses are agriculture and silviculture (Raulston et al. 1998). However, there is a potential of chemical pollution by pulp mills and waste products from cities and agricultural activities along tributaries in Georgia. For example, the City of Valdosta had major sewage spills of suspended solids into the Withlacoochee River from its wastewater treatment plant in March, July, and August 2013 (Enge et al. 2014). Although groundwater quality tends to be good in Florida, the middle reaches of the Suwannee River are increasingly being impacted by nutrients, particularly nitrates and nitrites, because of groundwater contamination from croplands, poultry farms, dairies, and septic tanks (Raulston et al. 1998; Knight 2015). Elevated nitrate concentrations can cause mortality of insects, larval fishes, and anurans (Kincheloe et al. 1979; Camargo and Ward 1995; Hecnar 1995), which are potential prey for Macrochelys, and possibly affect habitat suitability by increasing growth of filamentous macroalgae (Stevenson et al. 2007) and invasive hydrilla (Hydrilla verticillata; Kennedy et al. 2009).

Although forests bordering streams in the Suwannee drainage are mostly intact, residents often remove logs and fallen trees along the bank that obstruct boat traffic (Enge et al. 2014). In 2000, the State of Florida initiated a program to allow permitted removal of deadhead logs, which are submerged pine and cypress timber cut by axes up until the early 20th century (Kaeser and Litts 2008). From 2000 to 2008, more than 16,000 logs were removed from rivers, but this is likely a conservative estimate (Kaeser and Litts 2008). Deadhead logging is permitted in Florida in all of the Withlacoochee River, the farthest downstream portion of the Santa Fe River, and the Suwannee River from White Springs downstream to Fanning Springs. A logger reported locating several places on the Suwannee River with hundreds of logs (McFarland 2016). Removal of any woody debris from the Suwannee River and its tributaries could have a negative impact on Macrochelys, because woody debris may provide primary refugia during low-water periods and concentrate prey items (Enge et al. 2014).

In the species status assessment for M. suwanniensis, future conditions and viability of the species was projected in 50 yrs using a female-only, stage-structured matrix population model (USFWS 2020). Based on the current state of knowledge, M. suwanniensis populations are predicted to decline in abundance and range, but the current state of knowledge is full of uncertainty (USFWS 2020). This assessment would be strengthened with additional studies on population delineations, abundance and occupancy, variation in demographic rates across the range of the species, the impacts of threats on demography, and prevalence of threats across the landscape (USFWS 2020). Our findings were incorporated in this species status assessment. Population models based on our trapping data downstream of White Springs in the Suwannee River indicate a slightly decreasing or stable population (T.M.T., unpubl. data, 2021). Because of our trapping results, land-use patterns, and coverage of protected lands along the Suwannee and Santa Fe rivers, we are more optimistic regarding the future of M. suwanniensis populations in Florida than in Georgia. We believe that populations in Georgia are faring the best in the Alapaha drainage and in sections of the Little River and Okapilco Creek in the Withlacoochee drainage. Populations in Georgia streams that experienced commercial harvest in the past (Pritchard 1989) may continue to recover. Additional surveys of the Okefenokee Swamp and tributary streams of the upper Suwannee River in Georgia, most of which are difficult to access, are needed to better determine the status of populations at these sites. Our trapping results provide baseline data from which future population trends can be determined.