Use of Habanero Pepper Powder to Reduce Depredation of Loggerhead Sea Turtle Nests
Abstract
We tested the effectiveness of nest screens and habanero pepper powder as deterrents for reducing depredation of loggerhead sea turtle (Caretta caretta) nests on Sand Island, South Carolina, including nest-screening (1 × 1-m plastic or metal mesh secured over the nest; n = 33), surface-treatment (15 ml of habanero pepper powder sprinkled on the surface of the sand; n = 10), bottom-treatment (15 ml of habanero pepper powder sprinkled 3 cm above the top egg; n = 10), and 10 control nests. Twenty-two of 66 nests were partially or completely depredated by coyotes (Canis latrans), and nest survival rates varied from a high of 80% for surface-treated, 75% for screened, 45% for bottom-treated, and 32% for control nests. We propose that early surface-treatment of nests with pepper powder may provide an effective and more efficient management alternative to using screens, which are more difficult and costly to install, for reducing nest depredation at sea turtle nests.
Depredation of loggerhead sea turtle eggs and hatchlings occurs on almost all nesting beaches in the southeastern United States (National Marine Fisheries Service and US Fish and Wildlife Service [NMFS and USFWS] 2008). Accordingly, an objective of the Recovery Plan for the Northwest Atlantic Population of the Loggerhead Sea Turtle (Caretta caretta; NMFS and USFWS 2008) is to minimize annual mammalian nest-depredation rates within each recovery unit to ≤ 10% of nests. Common predators of loggerhead nests include armadillos (Dasypus novemcinctus), coyotes, feral hogs (Sus scrofa), foxes (Vulpes vulpes, Urocyon cinereoargenteus), ghost crabs (Ocypode quadrata), raccoons (Procyon lotor), and red fire ants (Solenopsis invicta; Stancyk 1982; Dodd 1988; Yerli et al. 1997; Engeman et al. 2003, 2005; Rusenko et al. 2007). In the absence of nest protection programs on some nesting beaches in the southeastern United States, mammals may partially or fully destroy up to 100% of all nests (Davis and Whiting 1977; Hopkins and Murphy 1982; Stancyk et al. 1980; Talbert et al. 1980; Schroeder 1981; Garmestani et al. 1997).
Several mammalian control methods are used at most major turtle nesting beaches in the southeastern United States to reduce depredation of sea turtle eggs and hatchlings (see NMFS and USFWS 2008 for review), including trapping and removal of predators (Engeman et al. 2005), flags (Longo et al. 2009), relocation of nests to fenced hatcheries (Stancyk et al. 1980; Talbert et al. 1980), conditioned taste aversion (Hopkins and Murphy 1982; Ratnaswamy et al. 1997), but most predominately screening of sea turtle nests (covering a nest with plastic or metal mesh or metal cage; Addison and Henricy 1994; Jordan 1994; Adamany et al. 1997; Ratnaswamy et al. 1997; Yerli et al. 1997; Baskale and Kaska 2005; Antworth et al. 2006; Kurz et al. 2011). At many nesting beaches, screening significantly improves hatching success by reducing nest depredation (Ratnaswamy et al. 1997; Antworth et al. 2006), but at other nesting beaches substantial nest loss still occurs despite the wide-scale use of screens and cages (Antworth et al. 2006). Miller et al. (2003) and Antworth et al. (2006) reported that screening can be less effective for loggerhead nests than nests of green (Chelonia mydas) and leatherback (Dermochelys coriacea) sea turtles because loggerhead nest depth is much shallower than nest depth of the latter species. Further, Mroziak et al. (2000) reported that nest caging might condition predators, such as raccoons, to the presence of a nest, thereby attracting predators. To reduce predator attraction at screened nests, Rusenko et al. (2007) reported significant decreases in raccoon and fox depredation of sea turtle nests when habanero pepper powder was used as a deterrent in conjunction with screens. Yet, it is uncertain whether the reduced depredation rates were a function of the pepper powder or the screens. Additionally, installing and maintaining screens is expensive and requires much effort (Ratnaswamy et al. 1997). Thus, the goals of this study were to test the differential effects of habanero pepper powder versus screening for reducing egg depredation at loggerhead sea turtle nests, and to evaluate its cost and ease of implementation.
METHODS
Study Site
The study was conducted on Sand Island, one of the 3 barrier islands of Yawkey Wildlife Center, a 9713-ha preserve located at the mouth of Winyah Bay in Georgetown County, South Carolina. Sand Island is a long (8 km), narrow, flat island with no human habitation and is only accessible by boat.
In summer 2010, loggerhead sea turtles nested along the outer beach from the southern tip extending north for 4.18 km. Mammal-track surveys conducted during summer 2010 indicated that coyotes, raccoons, and bobcats (Lynx rufus) occurred on the island (J. Sexton, unpubl. data, 2010), and ghost crabs were common. Prior to summer 2010, sea turtle monitoring and management on the island was minimal. In contrast, sea turtles are carefully monitored and managed on South Island located immediately south of Sand Island and separated by a narrow (47.8-m) inlet. In 2008 and 2009, coyotes were trapped extensively in an effort to reduce depredation on turtle eggs and hatchlings on South Island. Presumably, the predators on South Island also readily move to Sand Island. In 2010, 138 loggerhead sea turtle nests were found and screened on South Island.
Experimental Design
The study was conducted on Sand Island from 20 May to 30 September 2010. Loggerhead turtles lay their nests at night, but we searched for turtle nests during the day by following tracks to body pits. Eggs were located using a wooden or metal probe, and eggs were typically 30–60 cm below the surface of the sand. When found, sand on top of the egg clutches was temporarily removed to collect a single egg or eggshell fragment (for depredated nests) from each clutch for use in a separate phylogeography study of loggerhead turtles, and then refilled. A Global Positioning System coordinate was recorded for each nest, and a flag placed 1 m to the south of each nest.
Research permits dictated limited treatment replication. The first 7 nests found on the island at the beginning of the nesting season were screened using 1 × 1-m plastic and metal screens as described by Ratnaswamy et al. (1997). The next 30 nests laid were alternately assigned to 3 groups: control (n = 10), surface-treatment (n = 10), and bottom-treatment (n = 10). Neither screens nor pepper powder was applied to control nests. All subsequent turtle nests laid were screened (n = 26) except for 3 nests found late in the season (hereafter, termed “late find nests”).
For the surface-treatment, about 15 ml of habanero pepper powder made from red savina habanero peppers (Capsicum chinense; 400,000–500,000 Scoville units; Spontaneous Combustion, Southwest Specialty Foods, Inc) was sprinkled in a 0.5-m-diameter circle on the surface of the sand above the nest to intercept predators approaching the nest from various angles. A thin layer of sand was scattered on top of the powder to reduce the potential of disturbance by wind or rain. For the bottom-treatment, 15 ml of habanero chili powder was applied to the sand about 3 cm above the topmost egg in the nest, and the nest refilled with sand. Depth of nests (to the topmost eggs) typically ranged from 30 to 60 cm deep with diameters of about 30 cm wide at the bottom and 15 cm wide at the excavation hole at the surface.
The beach was monitored every 1–3 d from 5 June to 23 July with the exception of a 6-d period in mid-July when no surveys were conducted. During each nest visit, numbers of depredated and undisturbed nests were recorded, and at depredated nests, counts were made of numbers of intact and depredated eggs (estimated from shell fragments). Identification of nest predators was based on tracks found within 1 m around nests.
After 23 July, there were only 12 visits to Sand Island to check the status of nests. Considering all but 2 nests hatched after 23 July when regular monitoring ended, we lack specific hatch date information for nearly all turtle nests. The final nest inventory occurred on 30 September, when numbers of unhatched eggs and dead hatchlings were counted, and number of emerged hatchlings estimated from number of eggshell fragments at 40 nests. Because nests survived for various time lengths and were not monitored every day, we calculated number of nest-days surviving for each treatment following the basic approach of Mayfield (1961, 1975) and Trent and Rongstad (1974). Nest-days surviving were based on the number of days each nest was known to survive before the first depredation event, emergence of any hatchlings, or washed away by storms (Appendix 1). We considered a nest depredated if all or part of the clutch was lost, except for the partial clutch depredation by ghost crabs at 1 nest (SND001) where 15 eggs were lost. For calculating nest survival rates, we considered a nest successful if it had not been depredated, even if it failed to hatch. Relocated nests were not included in nest survival estimates. One nest (SND047) found depredated on 23 July late in the season was excluded from nest survival estimates because the day of loss was unknown. Loggerhead turtle egg incubation periods range from 46 to 82 d depending on incubation temperatures (Matsuzawa et al. 2002); we used an average of 75 d for calculating seasonal nest survival rates, representing the mode of 42 loggerhead nests that were not depredated in this study (X¯ = 69, SD = 10.58, min. = 42, max. = 87). Date of nest loss was assumed to be the midpoint between the last nest check when a nest was known to be undisturbed and the date when a nest was found depredated. Z-values were calculated using MICROMORT 1.3 (Heisey and Fuller 1985) and used to compare survival rates of loggerhead turtle nests by treatment type. We considered differences statistically different if p ≤ 0.1, considering the relatively small sample size of treated nests. A 2-sample t-test was used to compare numbers of eggs lost to coyotes at partially depredated screened and control nests.
RESULTS
Sixty-six loggerhead nests were laid on Sand Island during summer 2010, with the first nest laid on 20 May and the last on 3 August. Clutch size averaged 109.5 eggs (n = 40, SD = 16.2, range = 70–155 eggs). Peak turtle nesting occurred between 30 May and 26 June. Of the 66 nests monitored (including the 3 late find nests for which we were able to determine their fate), 22 nests were either completely or partially depredated. Control (n = 6 of 10) and bottom-treatment (n = 5 of 10) nests had the highest proportion of depredated nests, and screened (n = 7 of 33) and surface-treatment (n = 2 of 10) nests had the lowest. Nineteen nests were overwashed by high tides between 1 and 4 times during the nesting season without being washed away, and an additional 14 nests were completely washed away by storms between 7 and 28 August. However, overwashed and washed-away nests were kept in the analysis because they were available for depredation during the majority of incubation. Overall, nest survival rates were 1.7–2.5 times higher for surface-treatment (80%) and screened (75%) nests compared with bottom-treatment (45%) and control (32%) nests (Table 1). Differences in nest survival rates were greatest for surface-treatment (Z = 2.4, p = 0.016) and screened (Z = 2.54, p = 0.012) nests compared with control nests (Table 2). Nest survival rates were similar between surface-treatment and screened nests (Z = 0.29, p = 0.723), and bottom-treatment and control nests (Z = 0.59, p = 0.47; Table 2).
Coyotes were responsible for almost all depredated nests, except for one partially depredated nest (surface-treatment) with raccoon tracks, and a second partially depredated nest (bottom-treatment) with ghost crab tracks. Numbers of eggs lost in nests with partial depredation ranged from 13 to 127 (X¯ = 57.3, SD = 38.7, n = 9) for coyotes, and 12 and 63 eggs, respectively, for each of the nests depredated by ghost crabs and raccoons. There was no difference in average numbers of eggs lost for the 5 screened nests (X¯ = 63.4, SD = 46.48) and 3 control nests (X¯ = 55.0, SD = 36.01) partially depredated by coyotes (t = 0.265, P = 0.780). For all partially depredated nests, only nests with screens (n = 4) and one late find nest successfully hatched any turtle hatchlings. Although sample sizes were small for undisturbed nests that were also not overwashed, hatchlings emerged from all of late find (n = 1 of 1), surface-treatment (n = 1 of 1 nest), and bottom-treatment (n = 2 of 2 nests) nests and from 91% of screened nests (n = 10 of 11 nests).
The highest frequency (10 of 22 nests, 45%) of nest depredation by the 3 nest predators occurred within the first 2 d after eggs were laid (Fig. 1). However, coyotes continued to depredate nests throughout the incubation period, with a second peak between 61 and 80 d after egg-laying. For the 7 nests (1 surface-treatment, 1 control, 5 screened) relocated after partial depredation, no depredation occurred at 6 of these relocated nests. For the seventh relocated nest (SND065), 23 eggs from this partially depredated nest were transferred to another nest (SND066) that was completely depredated by coyotes 4 d after eggs were transferred.
Citation: Chelonian Conservation and Biology 12, 2; 10.2744/CCB-0945.1
DISCUSSION
Habanero pepper powder used as a surface-treatment at nests significantly increased the survival rate of loggerhead nests from depredation compared with control nests and bottom-treated nests. Similarly, screened nests also had significantly higher survival than control nests and bottom-treated nests, and were comparable to the survival rate for our surface-treated nests. Despite the weathering effects of wind and rain, the avoidance effect of the pepper powder at surface-treated nests appeared to persist during the early period of incubation—possibly up to 20 d—when depredation rates were highest (Fig. 1). Further, of the 2 surface-treated nests lost, 1 nest was depredated on day 1 after egg-laying and the second 88 d after egg-laying. We suspect that depredation at this later nest may have been associated with hatchling emergence, because depredation occurred long after the period when we would expect the powder to be effective. The high survival rate for surface-treated nests suggests that pepper powder can be an effective deterrent for reducing coyote depredation of sea turtle nests.
Our use of screens at nests was equally effective at reducing nest depredation as our surface pepper powder treatment, and several studies (Ratnaswamy et al. 1997; Yerli et al. 1997; Baskale and Kaska 2005; Antworth et al. 2006) reported reduced depredation at turtle nests using screens and/or cages. Yet, Mroziak et al. (2000) reported that screens attract raccoons, and depredation rates were higher at nests with screens compared with nonscreened nests. Further, installing and maintaining screens requires much effort, and costs of screening using volunteers in Florida averaged $7.52/nest (US dollars) for 3000 turtle nests in 1997 (Ratnaswamy et al. 1997). Cost for pepper powder was about $2.00/nest to treat 20 nests in our study.
In contrast to our surface-treatment, our bottom-treatment with pepper powder appeared to be less effective in reducing nest depredation rates. The factors affecting this relatively low nest survival rate for bottom-treated nests are unknown, but may be related to the absence of a strong cue at the surface from the buried pepper powder. In the absence of this cue, coyotes may start digging up the nest. Additionally, there may be a reduced concentration of the buried powder due to the mixing of sand with the powder when a nest is refilled or through the digging action of the coyote.
Although coyotes are considered “first night predators” (i.e., those that tend to depredate nests on the night of laying), only 3 of our 21 depredated nests were depredated on the first night eggs were laid. However, 12 of our nests were depredated within the first 20 d after egg-laying, indicating that nests were most vulnerable to predation during the early period of incubation. In contrast, first night depredation by coyotes was very common on South Island in 2010, where the beach was actively patrolled nightly for new nests to screen. The difference in these first night depredation rates may be related to the higher numbers of nesting turtles on South Island (n = 138) versus Sand Island (n = 66), and/or higher numbers of coyotes on South Island due to the absence of freshwater sources on Sand Island, which may limit coyote numbers there, and to the water inlet that coyotes must swim to move between the 2 islands. The second peak in coyote nest depredation 61–80 d after egg-laying was most likely associated with emergence of turtle hatchlings. Further, continued depredation of nests for several weeks after egg-laying at treated and control nests (n = 8 nests depredated between 6 and 60 d after egg-laying; Fig. 1) also suggest that the scent from egg-laying may persist for extended periods.
Despite the recent occurrence of coyotes on Sand Island in 2006, coyotes are now the most prevalent predator of loggerhead nests. Prior to coyotes occurring on South Island, raccoons accounted for all loggerhead sea turtle nest depredation there (Hopkins and Murphy 1982). Despite predator control in 2008 and 2009 on both South and Sand Islands, coyotes are well-established and replaced raccoons as the most frequent predator of turtle nests on South Island. In contrast, there are few reports of loggerhead nests depredated by coyotes at other beaches in the southeastern United States (Atencio 1994; Lewis et al. 1996; Northwest Florida Partnership 2000 as cited in Engeman et al. 2003). However, coyotes are expanding their distribution in the Southeast (Schrecengost et al. 2009) and wildlife managers are concerned that coyotes will become established at other beaches used by nesting sea turtles (K. Rusenko, pers. comm., November 2010).
Although based on only 1 season of data at 1 beach location in South Carolina, our study suggests that the exclusive use of habanero pepper powder at sea turtle nests appears to be effective for reducing nest depredation by coyotes, especially when used as a surface-treatment. However, we caution that the possible irritant effects of habanero pepper powder on turtle hatchlings are unknown, requiring further study. Although nest survival rates were similar for our surface-treated and screened nests, nest-screening may be less effective in the long term if predators learn to associate screens with turtle nests, as reported in Florida. Further, the costs and time for treating nests with pepper powder are less than for screening. Thus, early surface-treatment of nests with pepper powder may provide an important management alternative to using screens for reducing nest depredation at sea turtle nests, but we encourage further research on the possible irritant effects on hatchlings before widespread implementation. We also encourage further research into the combined effects of using screens in combination with pepper powder for reducing depredation of sea turtle nests.
Acknowledgments
We wish to thank Todd Fuller for assistance calculating nest survival estimates and students for their assistance in nest treatments, monitoring, and predator track surveys. The South Carolina Department of Natural Resources does not endorse this study.
Numbers of depredated loggerhead turtle nests by predator versus number of days after egg-laying during summer 2010, Sand Island, South Carolina, USA.
Contributor Notes
