Parasitism of box turtles (genus Terrapene) by botflies (Cistudinomyia [formerly Sarcophaga] cistudinis; Sarcophagidae) was first described by Packard (1882) on Terrapene carolina and McMullen (1940) on Terrapene ornata (Table 1). Most reports were anecdotal and pertained to T. carolina (e.g., review in Peters 1948), but 6 detailed the infestations on T. ornata (Table 1). This botfly was also reported to infest Gopherus polyphemus (Knipling 1937) and Chrysemys picta (Chidester 1915), but only once in each case. However, frequencies, patterns, and survival impacts of infestation on any of these species have not been evaluated.

Table 1. Literature records of botfly larvae in Terrapene ornata. CL = carapace length; NR = not recorded.

View Fullscreen

During my long-term study (1981–2018) of T. ornata on and near the Crescent Lake National Wildlife Refuge in the Sandhills in Garden County, western Nebraska, I captured 609 T. ornata a total of 2584 times. When processing each, I noted the presence of botflies, removed all botfly larvae from their vesicles using fine-tipped forceps, and then released the turtle (except in one case; Table 2).

Table 2. Details of botfly infestations on Terrapene ornata in western Nebraska. Estimated values for age were based on annulus counts along with recapture histories; other values were based on turtles aged precisely as small juveniles and subsequently recaptured. PL = plastron length; NR = not recorded; SA = subadult; J = juvenile.

View Fullscreen

Among all 609 captures, I observed botfly infestations on 12 turtles (2.0%), although it is possible that I missed the presence of some botflies on turtles that kept their shells tightly closed during processing (although I always opened the forelobe in order to measure plastron length in a flat plane). Infestation occurred in 7 females, 3 males, and 2 juveniles, at ages between 8 and at least 27 winters (age based on recaptures of juveniles aged by counts of plastral annuli). No other study has quantified rates of infestation in any turtle population.

No turtle was found to have botflies in more than 1 yr, and 7 of the 11 were not captured again (the 12th was dissected for reproductive data). The remaining 4 were recaptured during 2 to 4 subsequent years over the following 1 to 19 yrs (Table 2). The 7 that were never recaptured contained 2–28 larvae (mean 7.9), whereas 3 of the 4 that were subsequently recaptured bore only 1–4 larvae (mean 2.7) (larval numbers were not recorded for 1 recaptured turtle), suggesting that the likelihood of death may increase with the size of the infestation. However, the difference was not significantly different (Mann-Whitney U test; Z = 1.25; p = 0.21). Legler (1960) observed botflies to be more common in subadults than adults in Kansas; however, I found that only 4 of 12 infested turtles were subadults (Table 1).

Legler (1960) questioned whether botflies contributed significantly to mortality in Ornate Box Turtles. However, 7 of 11 turtles in my study were never recaptured again, and may have had higher parasite loads than those that were recaptured (Table 2). Annual survival in this population was 88% for males and 93% for females (Converse et al. 2005), and recapture rates were generally high. For example, 25 of 44 (57%) of box turtles captured in 1985 were recaptured in 1986, and 49 of 79 (62%) captured in 1998 were recaptured in 1999. These high rates of survival and recapture suggest that turtles may have died from the infestation even though I carefully removed the larvae.

Only 2 Ornate Box Turtle studies reported the fates of infested turtles. In one, the turtle died overnight (14 larvae present; Rainey 1953) and in the other, the turtle survived at least to the following spring after larval removal and medical treatment by a veterinarian (O’Toole et al. 2021). However, Peters (1948) reported the death of a T. carolina 8 d after he removed 24 botfly larvae, and a second turtle died 7 mo after 19 larvae were excised. Rokosky (1948) noted the death of a captive T. carolina (bauri) that was found during necroscopy to be infested with 23 botfly larvae.

Another T. carolina captured in the fall died in April the following spring, and upon its death was found to have 8 botfly larvae in its neck (Wheeler 1890). In addition, King and Griffo (1958) attributed the death of a Florida T. carolina (bauri) to the presence of 70 larvae. These data suggest that larger infestations of botflies may increase mortality in box turtles.

All but 1 of my 12 infested turtles were captured on the upland side of drift fences placed parallel to the shoreline of Gimlet Lake (i.e., headed toward the water) between 15 May and 8 June. This pattern suggests that they were dehydrated and in search of water. It is also possible that submergence in water could be a mechanism to shed larvae by limiting oxygen.

There are a few published accounts on the potential physical mechanics of botfly infestation leading to box turtle mortality in Terrapene. Peters (1948 for T. carolina) suggested that botfly myiasis might impede locomotion and result in starvation or desiccation. Chidester (1915) reported 5 botfly larvae in the thigh of a C. picta that caused the complete paralysis of the limb. The Ornate Box Turtle that died the day after capture by Rainey (1953) had an empty intestine, perhaps related to the infestation. Apart from these anecdotal observations, the physical and physiological impacts of botflies on box turtle hosts are completely unknown, as is the phenology of infestation.

Interestingly, infestation rates in my population seemed to vary by year, with no sightings in 9 full spring sample years (1984, 1985, 1986, 1988, 1990, 1993, 2006, 2007, 2017), 1 each in 1998 and 2015, 3 in 2014, and 7 in 1999. Hence, it is possible that botfly infestations are climate related. To test that hypothesis, I examined monthly temperature and rainfall data for the 10 mo prior to the spring sample year, and found that the only variable that was correlated with infestation frequency was the mean daily temperature in September in the year prior to capture (Fig. 1). Whether this relationship has biological significance for infestation remains to be studied, but the fall of 1998 seems to have presented optimal conditions for transmission and development into the following year.

View Figure

• Download as Powerpoint
• Download Figure

Wheeler (1890) received a female T. carolina in the fall, but it died the following April, suggesting that botflies may place eggs on turtles in the fall, and the larvae may overwinter in the turtle. Knipling (1937) also asserted that botfly larvae can overwinter on the host. Legler (1960), however, reported finding an infested female T. ornata in October in a hibernaculum, but the following June the turtle was no longer infested, suggesting that the larvae emerged before June.

These data suggest that botfly infestation, though uncommon, can often result in host mortality. Future studies should be directed to the mechanics of transmission to box turtles, the potential impact of climate on transmission, the actual pathology leading to death, and the comparative physiology of infested and uninfected turtles.

Adult botflies typically lay eggs near areas frequented by their hosts, and once the larvae hatch they crawl onto the host and permeate the skin, where they consume the flesh of the host (O’Toole et al. 2021). Wheeler (1890), Knipling (1937), McMullen (1940), and Peters (1948) each speculated that the eggs or young larvae may be deposited on any abraded area of the skin that facilitates penetration. Wheeler (1890) suggested that the flies may use old infestation sites for oviposition. However, I did not find a repeated infestation on turtles, although only 3 turtles were recaptured after botfly removal.

My observations of fully developed larvae were confined to springtime; I did not sample between mid-July and the following winter. Legler (1960) reported finding infected box turtles throughout the activity season, and captures of infected Terrapene were reported in late July (Emerton 1904), August (Peters 1948; Rodeck 1949; Rainey 1953), September (Peters 1948), and November (McMullen 1940; Rokosky 1948), corroborating Legler’s observations.

In conclusion, despite the many anecdotal reports of botfly infestations of box turtles, little is known about the physiological impacts of infestation, the phenology of infestations, or the cues used by botflies to select a host. However, the mortality of infested turtles is high, and may significantly impact population survival where botflies are prevalent.