The paleoecology and extinction of endemic tortoises in the Bahamian Archipelago

Abstract

No native species of tortoises (Chelonoidis spp.) live today in the Bahamian (Lucayan) Archipelago (= The Bahamas + The Turks and Caicos Islands), although a number of species inhabited these islands at the first human contact in the late-Holocene. Until their extinction, tortoises were the largest terrestrial herbivores in the island group. We report 16 accelerator mass spectrometer (AMS) radiocarbon (¹⁴C) dates determined directly on individual bones of indigenous, extinct tortoises from the six Bahamian islands (Abaco, Eleuthera, Flamingo Cay, Crooked, Middle Caicos, Grand Turk) on five different carbonate banks. These 16 specimens probably represent six or seven species of tortoises, although only one (Chelonoidis alburyorum on Abaco) has been described thus far. Tortoises seem to have survived on most Bahamian islands for only one or two centuries after initial human settlement, which took place no earlier than AD ~700–1000. The exception is Grand Turk, where we have evidence from the Coralie archeological site that tortoises survived for approximately three centuries after human arrival, based on stratigraphically associated ¹⁴C dates from both tortoise bones and wood charcoal. The stable isotope values of carbon (σ¹³C) and nitrogen (σ¹⁵N) of dated tortoise fossils show a NW-to-SE trend in the archipelago that may reflect increasing aridity and more consumption of cactus.

Keywords

human impact late-Holocene radiocarbon dates stable isotopes vertebrate fossils West Indian Islands

Introduction

Across the globe, vertebrates of all sorts have suffered numerous instances of late Quaternary extinction (species-level losses) and extirpation (losses of insular populations of a species that survives elsewhere). Bones of extinct or extirpated species can occur in both paleontological (non-cultural) and archeological (cultural) sites, the later establishing the former co-existence of people with lost species or populations of reptiles, birds, and mammals. Cultural sites also are more likely to contain charcoal or other organic materials suitable for radiocarbon dating. For non-cultural sites, absolute chronologies often depend on dating the bones themselves, if they retain adequate quantities of collagen.

Such has been the case in the West Indies, where the bones of extinct, endemic species of ground sloths, monkeys, large rodents, reptiles, and birds have been radiocarbon dated to the late-Holocene in Hispaniola, Cuba, and the Bahamas (Cooke et al., 2017; Steadman et al., 2005, 2007, 2015). Lying north of Cuba and Hispaniola, the Bahamian Archipelago (also called the Lucayan Archipelago) was the last West Indian island group colonized by Amerindians. From cal. AD ~700–1000 (= ~1250–950 cal. BP), people began to inhabit these low-lying limestone islands, which are surrounded by rich marine resources but with more limited and vulnerable terrestrial plant and animal communities (Berman and Gnivecki, 1995; Carlson, 1999; Carlson and Keegan, 2004; Keegan, 1997; Keegan et al., 2008; Newsom and Wing, 2004).

Unlike in the nearby Greater Antilles (Cuba, Jamaica, Hispaniola, Puerto Rico; Figure 1), the Bahamian islands lacked large mammals at human contact; the largest native Bahamian terrestrial mammal was the hutia Geocapromys ingrahami, a medium-sized rodent (0.4–0.8 kg body mass). By far, the largest indigenous Bahamian herbivores overall were tortoises (Chelonoidis spp.), none of which survives. Among these tortoises, only Chelonoidis alburyorum has been described thus far (Franz and Franz, 2009), although systematic studies of other Bahamian tortoises are under way (see below). Along with Cuban crocodiles (Crocodylus rhombifer), tortoises were the largest terrestrial animals available to the earliest human settlers on the Bahamian islands (Hastings et al., 2014; Morgan and Albury, 2013; Steadman et al., 2014).

Figure 1.

Map of the Bahamian Archipelago and adjacent Greater Antillean islands. Red dots indicate locations of fossil tortoises. (Only those from the Bahamian Archipelago are mentioned in this paper.) Light greenish blue shading represents additional land that was exposed during Quaternary glacial intervals.

Here, we will summarize the results of radiocarbon dates and associated stable isotope data for carbon and nitrogen, on indigenous Bahamian species of tortoises. If not for human impact, these tortoises would still be part of Bahamian terrestrial ecosystems, or at least they would have survived until after European contact (Carlson and Keegan, 2004; Fitzpatrick and Keegan, 2007). After documenting the chronology of their extinction, we will discuss briefly the possible ecological roles (as browsers and seed dispersers) that tortoises once had in these islands.

Study area, materials, and methods

The Bahamian Archipelago (= The Bahamas + The Turks and Caicos Islands) extends southeastward from near Florida toward eastern Cuba and Hispaniola (Figure 1). These islands lie on shallow-water carbonate banks that were fully exposed above the ocean during the late Pleistocene but are largely submerged today. The sea-level-controlled expansion and contraction of land on these banks through Quaternary glacial–interglacial cycles (Carew and Mylroie, 1995; Hearty et al., 1998; Milne and Peros, 2013) has strongly influenced the composition and distribution of the archipelago’s terrestrial plant and animal communities (Steadman and Franklin, 2015).

The arrival of humans to the Caribbean islands had profound impacts, including species translocation, extirpation, and extinction of various species, on terrestrial and marine animals (Giovas, 2019; Newsom and Wing, 2004; Steadman et al., 2017a; Steadman and Franklin, 2015; Wing, 2012). Present evidence indicates that the Bahama Archipelago was first inhabited by humans shortly after AD 700, with the oldest, securely dated human occupation at the Coralie site on Grand Turk (Carlson, 1999). Cultural materials in the site reflect the seasonal occupation of the island by foragers from Hispaniola who captured and consumed sea turtles (Chelonia mydas), iguanas (Cyclura carinata), bush- and ground-nesting birds, and tortoises (Carlson, 1999). The characteristics of these targeted prey types suggest that the site reflects the initial human presence on the island, where available animals with large body sizes or terrestrial habitats were the focus of early subsistence efforts. The prey types at the Coralie site, including tortoises, are largely absent from the later Lucayan sites.

Subsequent settlement of the rest of the Bahamian Archipelago took place about AD 1000, although the sequence and timing for most individual islands is not well established. A coherent cultural assemblage, called Lucayan after the Spanish designation of the Bahamas as ‘las islas de los Lucayos’ (Keegan and Carlson, 2008), existed by AD 1000. The Lucayans lived in small coastal communities that were periodically abandoned and reoccupied. Their material culture included locally made pottery vessels (Palmetto Ware), shell tools, and shell disk beads (Berman et al., 2013). They practiced a subsistence economy based on the slash-and-burn cultivation of maize (Zea mays) and manioc, harvesting of coontie (Zamia sp.), mollusk collecting, and fishing. Paleontological, archeological, and radiocarbon evidence indicates a strong correlation between Lucayan presence and changes in native animal biodiversity, including the translocation and extirpation of Bahamian hutias, the extirpation of iguanas and Cuban crocodiles, and the extinction of Creighton’s caracara (Caracara creightoni) and other birds (LeFebvre et al., 2019; Oswald et al., 2019; Steadman et al., 2014).

All radiocarbon dating and stable isotope analyses (¹³C/¹²C ratios, ¹⁵N/¹⁴N ratios) in this paper took place at Beta Analytic Inc. The calibration dataset follows IntCal13. Each age determination is an accelerator mass spectrometer (AMS) radiocarbon (¹⁴C) date on purified collagen from individual bones of tortoises of known provenience from paleontological or archeological sites. For either type of site, we purposely chose specimens for AMS ¹⁴C dating that appeared to be ‘young’ based on their provenience and/or physical preservation. Because of potential cultural associations and the predominance of reporting ¹⁴C dates in the Caribbean archeological literature as ‘cal. AD/BC’ rather than ‘cal. BP’, we use cal. AD/BC in the text, with values given in Table 1. Details on the methods of AMS ¹⁴C dating and stable isotope analyses, including pretreatment of the bones, are in https://www.radiocarbon.com/radiocarbon-dating-p.htm.

Table 1.

Radiocarbon (¹⁴C) and stable isotope (δ¹³C, δ¹⁵N) data for specimens of tortoise (Chelonoidis spp.) from the Bahamian Archipelago. C: cultural site; NC: non-cultural site; FS: field series (see Carlson, 1999); for MC-37, Unit: excavation unit, Roman numeral: stratigraphic layer, Arabic numeral: level (overall depth). NMB: National Museum of the Bahamas. Calibration follows IntCal13. Ages reported at 2σ (95.4% confidence). ^aSteadman et al. (2014). ^bData from Hastings et al. (2014). ^cTo calculate island mean values of δ¹³ C, we used the mean of the two reported values (generated in different labs) for specimens NMB.AB52.0039 and NMB.AB52.0040. ^dHerein. ^eSteadman et al. (2017a).

Bank and island	Site	Species	Skeletal element	Specimen	Lab number	Conventional ¹⁴C age (yr BP)	σ¹³C	¹⁴C age (cal. BP)	¹⁴C age (cal. AD/BC)	σ¹⁵N
Little Bahama Bank
Abaco^a	Gilpin Point (C)	C. alburyorum	Costal	NMB.AB62	Beta-338511	1010 ± 30	−21.6	960–910, 840–840	AD 990–1040, 1110–1110	+5.0
Abaco^b	Sawmill Sink (NC)	C. alburyorum	Humerus	NMB.AB50.0008	Beta-298219	940 ± 30	−20.3	970–920	AD 980–1030	+8.85
Abaco^b	Sawmill Sink (NC)	C. alburyorum	Scapula	NMB.AB50.0003	Beta-225509	2520 ± 50	−21.1	2770–2690, 2640–2610, 2590–2500	BC 820–740, 690–660, 640–550	−
Abaco^b	Sawmill Sink (NC)	C. alburyorum	Dorsal vertebra	NMB.AB50.0001	Beta-225508	2660 ± 40	−21.2	2880–2750	BC 930–800	−
Abaco^b	Lost Reel Cave (NC)	C. new sp. A	Carapace	NMB.AB52.0039	Beta-298220	880 ± 30	−22.1, −22.8^c	920–780	AD 1030–1170	+4.36
Abaco^b	Lost Reel Cave (NC)	C. new sp. A	Humerus	NMB.AB52.0040	Beta-298221	1130 ± 30	−20.9, −22.9^c	1230–1210, 1180–1060	AD 720–740, 770–890	+2.48
Great Bahama Bank
Eleuthera^d	Preacher’s Cave Blue Hole (NC)	Chelonoidis?	Costal	NMB.EL178	Beta-477771	400 ± 30	−12.6	514–428, 376–326	AD 1436–1522, 1574–1624	+7.7
Eleuthera^d	Kelly’s Blue Hole (NC)	C. new sp. B	Costal	NMB.EL.180.27	Beta-510248	No collagen	−	−	−	−
Jumentos Cays^d	Flamingo Cay Cave (C/NC?)	C. new sp. B	Plastral fragments	NMB.R11	Beta-510246	890 ± 30	−19.7	910–730	AD 1040–1220	+9.96
Crooked-Acklins Bank
Crooked^e	McKay’s Bluff Cave (C)	C. new sp. C	Fibula	NMB.CR05	Beta-451745	870 ± 30	−21.2	925–785	AD 1025–1165	+12.4
Crooked^e	1702 Cave (NC)	C. new sp. C	Costal	NMB.CR26	Beta-445995	2510 ± 30	−19.7	2740–2490	BC 790–540	+10.2
Mayaguana Bank
Mayaguana^d	Abraham’s Bay Cave (NC)	C. new sp. D	Plastral fragments	NMB.MY014.3	Beta-510247	No collagen	−	−	−	−
Caicos Bank
Middle Caicos^d	Indian Cave (NC)	C. new sp. E	Dentary	MC-37, Unit 9, surface	Beta-330406	1090 ± 30	−18.7	1060–930	AD 890–1020	+10.3
Middle Caicos^d	Indian Cave (NC)	C. new sp. E	Femur	MC-37, Unit 4, II/7	Beta-338517	1330 ± 30	−19.9	1300–1240, 1200–1180	AD 650–710, 750–770	−
Middle Caicos^d	Indian Cave (NC)	C. new sp. E	Costal	MC-37, Unit 4, II/4	Beta-338516	1890 ± 30	−19.9	1890–1700, 1760–1740	AD 60–180, 190–210	+9.9
Middle Caicos^d	Indian Cave (NC)	C. new sp. E	Plastron	MC-37, Unit 4, I/2	Beta-330407	1930 ± 30	−17.4	1940–1940, 1930–1820	AD 10–10, 20–130	+8.3
Middle Caicos^d	Indian Cave (NC)	C. new sp. E	Peripheral	MC-37, Unit 3, II/4	Beta-330408	No collagen	−	−	−	−
Middle Caicos^d	Indian Cave (NC)	C. new sp. E	Costal	MC-37, Unit 3, II/8	Beta-330409	No collagen	−	−	−	−
Middle Caicos^d	Indian Cave (NC)	C. new sp. E	Peripheral	MC-37, Unit 3, III/13	Beta-330410	No collagen	−	−	−	−
Middle Caicos^d	Indian Cave (NC)	C. new sp. E	Costal	MC-37, Unit 3, IV/17	Beta-330411	No collagen	−	−	−	−
Middle Caicos^d	Indian Cave (NC)	C. new sp. E	Femur	MC-37, Unit 4, II/4	Beta-338515	No collagen	−	−	−	−
Turks Bank
Grand Turk^d	Coralie (C)	C. new sp. F	Scapula	GT-3 FS 82	Beta-477772	880 ± 30	−14.3	903–846, 833–728	AD 1042–1104, 1117–1222	+11.44
Grand Turk^d	Coralie (C)	C. new sp. F	Pleural	GT-3 FS 234	Beta-477775	1170 ± 30	−14.4	1179–1047, 1032–985	AD 771–903, 918–965	+8.2
Grand Turk^d	Coralie (C)	C. new sp. F	Costal	GT-3 FS 343	Beta-477773	No collagen	−	−	−	−
Grand Turk^d	Coralie (C)	C. new sp. F	Two costals	GT-3 FS 348	Beta-477774	No collagen	−	−	−	−
Grand Turk^d	Coralie (C)	C. new sp. F	Pleural	GT-3 FS 236	Beta-477776	No collagen	−	−	−	−
Mean, range, and sample size for isotope values (does not include anomalous Beta-477771)							−19.6, −22.4 to −14.3, and 15		+6.8, +2.48 to +12.4, and 15

Our 16 direct AMS ¹⁴C dates on Bahamian tortoise bones are from six islands that span the archipelago; seven of these dates have been reported elsewhere (Hastings et al., 2014; Steadman et al., 2007, 2015); the other nine AMS ¹⁴C dates are presented here for the first time. This study had its origins with the discovery on Abaco and subsequent description of C. alburyorum, which is represented by the most abundant and high-quality fossils of any West Indian species of tortoise (Franz and Franz, 2009). The excellent preservation of collagen in the bones of C. alburyorum even permitted the extraction of ancient DNA, allowing its phylogenetic placement based on both morphological and molecular data (Kehlmaier et al., 2017).

We suspect that each carbonate bank in the archipelago harbored at least one endemic species of tortoise, although the fossils now available from some islands are not diagnostic enough to substantiate this claim. Nevertheless, we note here that diagnostic tortoise fossils from several other Bahamian islands are under study now by Franz and Albury; we foresee future taxonomic descriptions for additional species from Abaco (C. new species A), Long Island (C. new species B), Middle Caicos (C. new species E), and Grand Turk (C. new species F).

Most of the AMS ¹⁴C dates that we report here are from sites already mentioned in the literature, including Gilpin Point and Sawmill Sink on Abaco (Franz and Franz, 2009; Steadman et al., 2007, 2014, 2015), McKay’s Bluff Cave and 1702 Cave on Crooked Island (Steadman et al., 2017a), Indian Cave on Middle Caicos (Carlson et al., 2006; Franz et al., 2000), and the Coralie site on Grand Turk (Carlson, 1999; Carlson et al., 2006; Carlson and Keegan, 2004; Franz et al., 2000; Keegan et al., 2008). Five previously unreported sites each have yielded a single AMS ¹⁴C determination on tortoise bone. They are Lost Reel Cave, a flooded site on Abaco; Preacher’s Cave Blue Hole and Kelly’s Blue Hole, two flooded sites on Eleuthera; Flamingo Cay Cave, a fairly dry site near the sea level on Flamingo Cay in the Jumentos; and Abraham’s Bay Cave, a dry cave on Mayaguana. (Precise site locations are not reported to prevent looting but can be requested from the Antiquities, Monuments and Museum Corporation in Nassau, The Bahamas.)

Results

Radiocarbon dates

All AMS ¹⁴C dates on individual bones of tortoises are within the late-Holocene (⩽3000 cal. BP; Table 1). On Abaco, where we have the most age determinations, the youngest dates on both species of tortoise (C. alburyorum and C. new species A) post-date cal. AD 1000, and therefore are within the period of human occupation (Figure 2). The dated specimen of C. alburyorum from Abaco’s Gilpin Point site (Beta-338511) is from a cultural context. Similarly, on Crooked Island, the youngest date on C. new species C (Beta-451745) is from a cultural site.

Figure 2.

Values of youngest AMS ¹⁴C dates (2σ or 95.4% confidence) on tortoise bone from six extinct species in the Bahamian Archipelago. Precise values for each vertical line are in Table 1. The horizontal dashed lines indicate the approximate time of human arrival for each island.

For islands on the Great Bahama Bank, the chronology of extinct tortoises is not well developed. We have not attempted to date any of the fragmentary tortoise fossils from the non-cultural Banana Hole site on New Providence, where the limited chronological evidence points to a late Pleistocene age (Morgan, 1989; Oswald and Steadman, 2018; Pregill, 1982). We also have not tried to date the tortoise fossils from the non-cultural Hanging Garden Cave site that we excavated on Long Island in December 2017; our attempt to obtain AMS ¹⁴C dates on the abundant fossils of hutia (G. ingrahami) from throughout the sediment profile of the Hanging Garden Cave was not successful because all six bones lacked collagen. We presume that site is the late Pleistocene rather than Holocene. Only two tortoise fossils from the Great Bahama Bank have yielded finite AMS ¹⁴C dates (Table 1). In an equivocally cultural setting on Flamingo Cay in the Jumento Cays, plastral fragments from a presumably undescribed form of tortoise yielded a date (Beta-510246; cal. AD 1040–1220) that is in close agreement with the youngest dates from Abaco and Crooked Island (Figure 2). A Lucayan skeleton found in Flamingo Cay Cave is not clearly associated with the tortoise bones.

On the other hand, the very young date (Beta-477771) from an undiagnostic tortoise costal bone on Eleuthera (cal. AD 1436–1522, cal. AD 1574–1624) overlaps with historic times (post-1492 AD) and may not represent an indigenous tortoise. European and American sailors often carried giant tortoises as food aboard ships (Townsend, 1925), opening up the possibility that the specimen from Eleuthera originated somewhere else. This possibility is strengthened by the unusual σ¹³C value of this specimen (−12.6‰), which is the least negative among all dated Bahamian tortoises and suggests a non-indigenous diet with substantial quantities of marine-derived food, C₄ grasses, or crassulacean acid metabolism (CAM) plants. Captive tortoises can be highly opportunistic omnivores, thereby opening up a range of possibilities for their isotopic signatures.

Our attempt to date an undescribed tortoise from a non-cultural context on Mayaguana (C. new sp. D) was not successful. From surface remains (Beta-330406) and non-cultural excavations (Beta-330407, 338516, 338517) at the Indian Cave site (MC-37) on Middle Caicos Island, we obtained four late-Holocene AMS ¹⁴C dates on individual tortoises bones (C. new sp. E), with the youngest date similar in age to the youngest ones from Abaco, Jumentos Cays, Crooked Island, and Grand Turk (Table 1; Figure 2). Specimens from deeper in the excavations at MC-37 lacked collagen and could not be AMS ¹⁴C dated.

The Coralie archeological site (GT-3) on Grand Turk Island presents an unusual situation where we can compare the two AMS ¹⁴C dates on individual tortoise bones directly with stratigraphically associated conventional ¹⁴C dates on wood charcoal. Two of the GT-3 tortoise bones submitted for dating had inadequate collagen (FS 343, 146N102E, stratum 2, 49-64 cmbs, tortoise K; and FS 348, 146N102E, stratum 1 & 2, 39-64 cmbs, tortoise K; Table 1). The younger of the two tortoise AMS ¹⁴C dates (Beta-477772; cal. AD 1042–1104, cal. AD 1117–1222; FS 82; 110N107E; stratum 1; 37 cmbs) is in close agreement with charcoal dates from similar proveniences (Figure 3). The older tortoise AMS ¹⁴C date (Beta-477775; cal. AD 771–903, cal. AD 918–965; FS 234 – Level 5.8–6.2) also agrees with wood charcoal dates from similar proveniences (Figure 3). Together, these sets of dates argue for a longer interval of contemporaneity of people and tortoises on Grand Turk than on other islands in the archipelago.

Figure 3.

Values of ¹⁴C dates (2σ or 95.4% confidence) on tortoise bone and wood charcoal from early and late contexts at the Coralie site (GT-3), Grand Turk. The oldest AMS ¹⁴C date on tortoise bone is cal. AD 771–903, cal. AD 918–965 (Beta-477775). The oldest associated conventional charcoal-based ¹⁴C dates are cal. AD 650–885, cal. AD 665–905, and cal. AD 670–970 (Beta-80911, −93912, −98698). The youngest AMS ¹⁴C date on tortoise bone is cal. AD 1042–1104 and cal. AD 1117–1222 (Beta-477772). The youngest associated conventional charcoal-based ¹⁴C dates are cal. AD 970–1165, cal. AD 895–1145, and cal. AD 1040–1215 (Beta-98697, −93913, −98699). Conventional dates are from Carlson (1999) and Carlson and Keegan (2004). The horizontal dashed line indicates the approximate time of human arrival on Grand Turk.

Stable isotopes

Ignoring the anomalous value from Eleuthera (see explanation above), the stable isotope values (¹³C/¹²C or σ¹³C) for carbon from tortoise collagen vary from −22.4‰ to −14.3‰ (mean −19.6‰; Table 1). All but two of the 15 σ¹³C values range from −22.4‰ to −17.4‰ (mean −20.6‰), which are reasonable values for terrestrial browsing herbivores, such as tortoises, with a diet dominated by C₃ plants (Hastings et al., 2014).

There is a clear trend for the σ¹³C values to become less negative from northwest to southeast in the island group, ranging from −22.4‰ to −20.3‰ on Abaco, −21.2‰ to −19.7‰ on Crooked Island, −19.9‰ to −17.4‰ on Middle Caicos, and −14.4‰ to −14.3‰ on Grand Turk (Table 1). These data possibly suggest a progressively larger marine-derived component (such as from seabird guano; Anderson and Polis, 1999; Steadman, 2006) in the prehistoric terrestrial food web as one proceeds southeasterly in the island group. Seabirds such as boobies, shearwaters, terns, and so on feed on marine fish and invertebrates but nest on land. The bird bones most commonly recovered from site GT-3 on Grand Turk were from boobies (Sula spp.; Carlson, 1999). Island size also could contribute to the trend, with Abaco (1214 km²) being much larger than the southern islands (Crooked 258 km², Middle Caicos 190 km², Grand Turk 20 km², and Flamingo Cay 0.9 km²). We calculated land areas using the following link: https://www.daftlogic.com/projects-google-maps-area-calculator-tool.htm#

Another possible factor might be the tortoises on more southerly islands fed more frequently on seashells (fossil and modern) and other calcareous stones to develop gastroliths to aid in processing food, as reported for gopher tortoises (Gopherus polyphemus) in southern Florida (Moore and Dornburg, 2014).

Yet another explanation, perhaps more plausible than the others, is that the tortoise diet on southern islands, which are drier than the northern islands, included a greater contribution from cacti. Many arid-adapted plants, including cacti, use CAM during photosynthesis. This pathway produces carbon isotope ratios that are in the range associated with C₄ species. The edible portions of three prickly pear cactus (Opuntia sp.) from Grand Turk yielded σ¹³C values of −10.2‰, −10.6‰, and −10.69‰ (Pestle, 2010). Prickly pear and Turk’s head cactus (Echinocactus horizonthalonius) are common on the island, the former nearly covering the GT-3 site prior to land clearance. Cacti (Opuntia sp.) are a favored food of tortoises in the Galápagos Islands (Blake et al., 2012; Gibbs et al., 2010).

Finally, we note that tropical grasses and sedges (C₄ plants) are an important dietary component for giant tortoises (Aldabrachelys gigantea) on Aldabra Atoll in the Indian Ocean (Walton et al., 2019). The inter-island distribution of C₃ vs C₄ grasses and sedges in the Bahamian Archipelago is not known well enough to evaluate how they may have affected the diet of extinct tortoises.

For nitrogen isotopes (¹⁵N/¹⁴N, or σ¹⁵N), the four tortoise specimens from Abaco are much less positive (+2.48‰ to +8.85‰; mean +5.17‰) than those from Crooked Island (+10.2‰ to +12.4‰; mean +11.3‰), Middle Caicos (+8.3‰ to +10.3‰; mean +9.5‰), or Grand Turk (+8.2‰ to +11.44‰; mean +9.82‰; Table 1). These values need to be reduced by ~2‰ to convert collagen values to the values of the consumed plants (Keegan and DeNiro, 1988). For the West Indies, Pestle (2010) reported a σ¹⁵N range for most terrestrial plants of +2‰ to +6‰, with Opuntia having a relatively high value (+6‰ to +7‰). The more enriched values in the southern islands can be explained in two related ways: first would be the greater dietary contribution of CAM plants (Opuntia) and second would be the high σ¹⁵N values reported generally for herbivores in areas with <400-mm annual rainfall (Sealy et al., 1987). The σ¹⁵N values in the Bahamas are similar for both terrestrial and coral reef animals (+5‰ to +10‰; Keegan and DeNiro, 1988), so either of these hypotheses could apply.

Discussion

Chronology of extinction

Late-Holocene extinctions and extirpations in the Bahamian Archipelago involved reptiles, birds, and mammals of all available sizes, ranging from small lizards to crocodiles, songbirds to caracaras, and bats to hutias (Franklin and Steadman, 2015; Soto-Centeno and Steadman, 2015). In some instances, an anthropogenic cause is highly likely, whereas in other cases, it is suspected although causation in less certain. Being large, easy to hunt, and highly edible, the loss of tortoises seems likely to be human caused.

Across the Bahamian Archipelago (Figure 1), the extinction chronology of tortoises shows little variation (Figure 2). On Abaco, for example, the youngest of four AMS ¹⁴C dates on C. alburyorum is cal. AD 990–1040, cal. AD 1110, whereas the youngest of two dates on C. new sp. A is cal. AD 1030–1170. Based on the youngest AMS ¹⁴C dates currently available, additional species of Chelonoidis survived elsewhere to roughly similar times on Flamingo Cay (cal. AD 1040–1220), Crooked Island (cal. AD 1025–1165; two dates), Middle Caicos (cal. AD 890–1020; four dates), and Grand Turk (cal. AD 1047–1104, cal. AD 1117–1222; two dates). Of course, we can never be sure that the 2σ age range of the youngest AMS ¹⁴C date on any given island captures the actual time of extinction. Nevertheless, it seems likely that tortoises typically were lost within a century or two of initial human arrival.

An exception is the Coralie site (GT-3) on Grand Turk. The three oldest conventional charcoal-based ¹⁴C dates from GT-3 (cal. AD 650–885, cal. AD 665–905, cal. AD 670–970; Beta-80911, −93912, −98698; Carlson, 1999; Carlson and Keegan, 2004) broadly overlap with our oldest AMS ¹⁴C date on associated tortoise bone (cal. AD 771–903, cal. AD 918–965; Figure 3 herein). Similarly, the three youngest conventional charcoal-based ¹⁴C dates from GT-3 (cal. AD 970–1165, cal. AD 895–1145, cal. AD 1040–1215; Beta-98697, −93913, −98699; Carlson, 1999; Carlson and Keegan, 2004) broadly overlap with our youngest AMS ¹⁴C date on associated tortoise bone (cal. AD 1042–1104, cal. AD 1117–1222; Figure 3 herein). A possible explanation for this relatively long period of co-existence of people and tortoises on Grand Turk might be that the older parts of GT-3 represent temporary encampments of people from the much larger and longer inhabited Hispaniola (165 km south of Grand Turk), rather than long-term settlement.

The generally rapid extinction of Bahamian tortoises stands in contrast to the protracted period during which endemic capromyid rodents (the hutia G. ingrahami) co-existed with people. Evidence for this late survival includes an AMS ¹⁴C date on a hutia bone from Garden Cave, Eleuthera (cal. AD 1500–1660; Beta-330400), and four hutia-based AMS ¹⁴C dates from the 15th–17th centuries AD from the smaller and more isolated Crooked Island (Steadman et al., 2017a, 2017b). On both islands, therefore, hutias survived into historic times (= post-AD 1492), a trend that is likely to be the case on other islands as well. While the hutia was also a popular food item prehistorically, this rodent undoubtedly had a much larger population and much shorter generation time than the tortoises and thus was able to withstand predation from Lucayans. That the extirpation of Bahamian hutia populations was post-Columbian suggests that Old World pathogens, perhaps from non-native rodents (Rattus spp.), were involved in their demise.

Island tortoises in paleoecological perspective

Tortoises in the Galápagos Islands are effective seed dispersers for a broad range of woody and herbaceous plants (Blake et al., 2012) and may serve as ‘ecological engineers’ in controlling the density and size classes of Opuntia cactus (Gibbs et al., 2010). Abundant fruits of mastic (Mastichodendron foetidissimum and satinleaf Chrysophyllum oliviforme; both Sapotaceae) were found inside the late-Holocene fossil carapaces of C. alburyorum from Sawmill Sink, Abaco (Franz and Franz, 2009). The Bahamian broadleaf forests are rich sources of these and many other types of fallen fruits as well as cactus of several genera, including Opuntia (Franklin et al., 2015; Franklin and Steadman, 2015). We believe that Bahamian tortoises once played a major role in seed dispersal, likely similar to that reported for living congeners in the Galápagos Islands (Hansen et al., 2010).

Knowing that various species of tortoises once inhabited essentially the entire Bahamian Archipelago, we encourage consideration of a program to translocate living, congeneric species of tortoises to selected islands. Such a program might re-establish the ecological equivalent of what once was the largest browsing herbivore and seed disperser on these islands. Possible candidate species of Chelonoidis consist of three continental forms (the red-footed tortoise C. carbonarius, yellow-footed tortoise C. denticulatus, and Chilean tortoise C. chilensis of mainland South America) and the Galápagos tortoise species group (C. niger s.l.; see Olson, 2017; Olson and David, 2014; Olson and Humphrey, 2017). With tortoises on tropical islands regarded as ‘ecological and evolutionary keystone species’ (Hansen et al., 2010), introducing non-native extant tortoises as replacements for extinct species is a reasonable and feasible ecological restoration scheme (Griffiths et al., 2010). After due consideration, any of the living species of Chelonoidis might have potential for introduction to selected Bahamian islands. Such efforts may restore some of the processes under which the plants of these islands evolved.

Postscript

On 1–2 September 2019, while we were addressing the reviewers’ and editor’s comments on this manuscript, Hurricane Dorian devastated Marsh Harbour and nearby cays on Abaco, causing massive loss of life and property. Co-authors Nancy Albury and Brian Kakuk survived the terrible storm, although they both lost essentially all personal and professional infrastructure (homes, offices, laboratories, equipment, vehicles, and so on). In spite of damage to a number of tortoise fossils in the NMB Collection in Marsh Harbour, most of them withstood the hurricane. They now reside at the Florida Museum of Natural History, where restoration of the collection has begun.

Footnotes

Acknowledgements

For permission to conduct the research and related courtesies in the Bahamas, we thank the National Museum of the Bahamas/Antiquities, Monuments and Museum Corporation, Abaco Friends of the Environment, and Bahamas National Trust (Michael Albury, Ruth Albury, Cha Boyce, Eric Carey, Michael Pateman, Olivia Patterson, Adriana McPhee Swain, Keith Tinker, and Kristin Williams). For permission to conduct the research and related courtesies in the Turks & Caicos Islands, we thank Brian Riggs. We appreciate the efforts of colleagues, namely, Kenny Broad, Nicole Canarrozzi, Kelly Delancy, Neil Duncan, Janet Franklin, Harlan Gough, Sharyn Jones, Michael Lace, Patrick O’Day, Jessica Oswald, Hayley Singleton, Angelo Soto-Centeno, Anne Stokes, Oona Takano, and Neill Wallis, who have worked with us in the field or laboratory. For comments that improved the manuscript considerably, we thank Janet Franklin.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study received funding from the National Science Foundation (BCS-1118340, BCS-1118369, and GSS-1461496) and the UF Ornithology Endowment.

ORCID iD

David W Steadman

References

Anderson

Polis

(1999) Nutrient fluxes from water to land: Seabirds affect plant nutrient status on Gulf of California islands. Oecologia 118: 324–332.

Berman

Gnivecki

(1995) The colonization of the Bahama Archipelago. World Archaeology 26: 421–441.

Berman

Gnivecki

Pateman

(2013) The Bahama Archipelago. In: Keegan

Hofman

Rodriguez

(eds) The Oxford Handbook of Caribbean Archaeology. New York: Oxford University Press, pp. 264–280.

Blake

Wikelski

Cabrera

et al. (2012) Seed dispersal by Galápagos tortoises. Journal of Biogeography 39: 1961–1972.

Carew

Mylroie

(1995) Quaternary tectonic stability of the Bahamian Archipelago: Evidence from fossil coral reefs and flank margin caves. Quaternary Science Reviews 14: 144–153.

Carlson

(1999) Aftermath of a feast: Human colonization of the southern Bahamian Archipelago and its effects of the indigenous fauna. PhD Dissertation, University of Florida, Gainesville, FL.

Carlson

Keegan

(2004) Resource depletion in the prehistoric northern West Indies. In: Fitzpatrick

(ed.) Voyages of Discovery: The Archaeology of Islands. Westport, CT: Praeger Publishers, pp. 85–107.

Carlson

Steadman

Keegan

(2006) Birdland. Times of the Islands 74: 64–69.

Cooke

Mychajliw

Southon

et al. (2017) The extinction of Xenothrix mcgregori, Jamaica’s last monkey. Journal of Mammalogy 98: 937–949.

10.

Fitzpatrick

Keegan

(2007) Human impacts and adaptations in the Caribbean Islands: An historical ecology approach. Earth and Environmental Science Transactions of the Royal Society of Edinburgh 98: 29–45.

11.

Franklin

Steadman

(2015) Winter bird communities in pine woodland vs. broadleaf forest on Abaco, The Bahamas. Caribbean Naturalist 3: 1–18.

12.

Franklin

Ripplinger

Marcano-Vega

et al. (2015) Regional variation in Caribbean dry forest tree species composition. Plant Ecology 216: 873–886.

13.

Franz

(2009) A new fossil land tortoise in the genus Chelonoidis (Testudines: Testudinidae) from the northern Bahamas, with an osteological assessment of other Neotropical tortoises. Bulletin of the Florida Museum of Natural History 49: 1–44.

14.

Franz

Carlson

Owen

et al. (2000) Fossil tortoises from the Turks and Caicos Islands, BWI. In: Proceedings of the Bahamas Field Station (Symposium, March 1999), San Salvador, Bahamas, pp. 5–9.

15.

Gibbs

Sterling

Zabala

(2010) Giant tortoises as ecological engineers: A long-term quasi-experiment in the Galápagos Islands. Biotropica 42: 208–214.

16.

Giovas

(2019) The beasts at large – Perennial questions and new paradigms for Caribbean translocation research. Part I: Ethnozoogeography of mammals. Environmental Archaeology 24(2): 182–198.

17.

Griffiths

Jones

Hansen

et al. (2010) The use of extant non-indigenous tortoises as restoration tool to replace extinct ecosystem engineers. Restoration Ecology 18: 1–7.

18.

Hansen

Donlan

Griffiths

et al. (2010) Ecological history and latent conservation potential: Large and giant tortoises as a model for taxon substitutions. Ecography 33: 272–284.

19.

Hastings

Krigbaum

Steadman

et al. (2014) Domination by reptiles in a terrestrial food web of The Bahamas prior to human occupation. Journal of Herpetology 48: 380–388.

20.

Hearty

Neumann

Kaufman

(1998) Chevron ridges and runup deposits in the Bahamas from storms late in oxygen-isotope substage 5e. Quaternary Research 50: 309–322.

21.

Keegan

(1997) Bahamian Archaeology. Nassau, The Bahamas: Media Publishing.

22.

Keegan

Carlson

(2008) Mother sea turtle. Times of the Islands 82: 45–50.

23.

Keegan

DeNiro

(1988) Stable carbon- and nitrogen-isotope ratios of bone collagen used to study coral-reef and terrestrial components of prehistoric Bahamian diet. American Antiquity 53: 320–336.

24.

Keegan

Fitzpatrick

Sullivan-Sealy

et al. (2008) The role of small islands in marine subsistence strategies: Case studies from the Caribbean. Human Ecology 36: 635–654.

25.

Kehlmaier

Barlow

Hastings

et al. (2017) Tropical ancient DNA reveals relationships of the extinct Bahamian giant tortoise Chelonoidis alburyorum. Proceedings of the Royal Society B 284: 20162235.

26.

LeFebvre

DuChemin

deFrance

et al. (2019) Bahamian hutia (Geocapromys ingrahami) in the Lucayan realm: Pre-Columbian exploitation and translocation. Environmental Archaeology 24(2): 115–131.

27.

Milne

Peros

(2013) Data-model comparison of Holocene sea-level change in the circum-Caribbean region. Global and Planetary Change 107: 119–131.

28.

Moore

Dornburg

(2014) Ingestion of fossil seashells, stones and small mammal bones by gravid gopher tortoises (Gopherus polyphemus) in South Florida. Bulletin of the Peabody Museum of Natural History 55(1): 55–63.

29.

Morgan

(1989) Fossil Chiroptera and Rodentia from the Bahamas, and the historical biogeography of the Bahamian mammal fauna. In: Woods

(ed.) Biogeography of the West Indies: Past, Present, and Future. Gainesville, FL: Sandhill Crane Press, pp. 685–740.

30.

Morgan

Albury

(2013) The Cuban crocodile (Crocodylus rhombifer) from late Quaternary fossil deposits in The Bahamas and Cayman Islands. Bulletin of the Florida Museum of Natural History 52: 161–236.

31.

Newsom

Wing

(2004) On Land and Sea: Native American Uses of Biological Resources in the West Indies. Tuscaloosa, AL: University of Alabama Press.

32.

Olson

(2017) The early scientific history of Galapagos tortoises. Archives of Natural History 44: 241–258.

33.

Olson

David

(2014) The gender of the tortoise genus Chelonoidis Fitzinger, 1835 (Testudines: Testudinidae). Proceedings of the Biological Society of Washington 126: 393–394.

34.

Olson

Humphrey

(2017) The island of origin of Richard Harlan’s Galapagos tortoise Testudo elephantopus. Archives of Natural History 44: 110–117.

35.

Oswald

Allen

Witt

et al. (2019) Ancient DNA from a 2,500-year-old Caribbean fossil places an extinct bird (Caracara creightoni) in a phylogenetic context. Molecular Phylogenetics and Evolution 140: 106576.

36.

Oswald

Steadman

(2018) The late Pleistocene bird community of New Providence, Bahamas. The Auk: Ornithological Advances 135: 359–377.

37.

Pestle

(2010) Diet and society in prehistoric Puerto Rico: An isotopic approach. PhD Dissertation, Department of Anthropology, The University of Illinois at Chicago, Chicago, IL.

38.

Pregill

(1982) Fossil amphibians and reptiles from New Providence Island, Bahamas. Smithsonian Contributions to Paleobiology 48: 8–21.

39.

Sealy

Van Der Merwe

Lee Thorp

et al. (1987) Nitrogen isotopic ecology in southern Africa: Implications for environmental and dietary tracing. Geochimica et Cosmochimica Acta 51: 2707–2717.

40.

Soto-Centeno

Steadman

(2015) Fossils reject climate change as the cause of extinction of Caribbean bats. Scientific Reports 5: 7971.

41.

Steadman

(2006) Extinction and Biogeography of Tropical Pacific Birds. Chicago, IL: University of Chicago Press.

42.

Steadman

Albury

Kakuk

et al. (2015) The vertebrate community on an ice-age Caribbean island. Proceedings of the National Academy of Sciences of the United States of America 112: E5963–E5971.

43.

Steadman

Albury

Maillis

et al. (2014) Late-Holocene faunal and landscape change in The Bahamas. The Holocene 24: 220–223.

44.

Steadman

Albury

Mead

et al. (2017b) Holocene vertebrates from a dry cave on Eleuthera Island, Commonwealth of The Bahamas. The Holocene 28: 806–813.

45.

Steadman

Franklin

(2015) Changes in an insular bird community since the late Pleistocene. Journal of Biogeography 42: 426–438.

46.

Steadman

Franz

Morgan

et al. (2007) Exceptionally well preserved late Quaternary plant and vertebrate fossils from a blue hole on Abaco, The Bahamas. Proceedings of the National Academy of Sciences of the United States of America 104: 19897–19902.

47.

Steadman

Martin

MacPhee

RDE

et al. (2005) Asynchronous extinction of late Quaternary sloths on continents and islands. Proceedings of the National Academy of Sciences of the United States of America 102: 11763–11768.

48.

Steadman

Singleton

Delancy

et al. (2017a) Late- Holocene historical ecology: The timing of vertebrate extirpation on Crooked Island, commonwealth of The Bahamas. Journal of Island and Coastal Archaeology 12: 572–584.

49.

Townsend

(1925) The Galapagos tortoises in relation to the whaling industry. Zoologica 4: 55–135.

50.

Walton

Baxter

Bunbury

et al. (2019) In the land of giants: Habitat use and selection of the Aldabra giant tortoise on Aldabra Atoll. Biodiversity and Conservation 28: 3183–3198.

51.

Wing

(2012) Zooarchaeology of West Indian land mammals. In: Woods

Sergile

(eds) Terrestrial Mammals of the West Indies. Gainesville, FL: Florida Museum of Natural History, pp. 342–356.