Determining how long turtles live in the wild can be controversial because age estimates are often based on counts of growth rings, with no conclusive evidence that a growth ring is added each year (Wilson et al. 2003). However, plastral rings can approximate age until 15–17 yrs old, at which point abrasion usually obscures grooves (Germano and Bury 1998; Quinn et al. 2014; Bernstein et al. 2018).

There are several seemingly credible reports of eastern box turtles (Terrapene carolina) living over 100 yrs in the wild (Dodd 2001; Currylow et al. 2011). Legler (1960) speculated that ornate box turtles (T. ornata) could live over 50 yrs in Kansas, USA, but most past studies suggested a shorter lifespan (Edmonds 2020). Metcalf and Metcalf (1985) observed individuals at least 28 yrs old in Kansas and concluded that few, if any, turtles survived past 30 yrs. In support, Blair (1976) found 2 31-yr-old males and a 32-yr-old female in Texas, USA, and concluded that the population completely turned over every 32 yrs. However, Christiansen et al. (2004) documented a wild female in southeastern Iowa, USA, that was at least 37 yrs old and Germano (2014) estimated that 3 individuals lived over 40 yrs in New Mexico, USA. More recently, Lewis and Iverson (2018) presented data indicating that some turtles lived over 50 yrs and suggested that some could even survive beyond 60 in Nebraska, USA.

Maximum lifespan influences important demographic processes and patterns for species such as survivorship, individual reproductive potential, recruitment, and population viability (Congdon et al. 2003; Kuo and Janzen 2004; Warner et al. 2016; Reinke et al. 2020). Empirical knowledge of longevity in the wild can inform management decisions by the incorporation of accurate ages into population viability analyses (Brook et al. 2000). Age itself can also relate to growth rate, which is useful in intraspecies comparisons to understand the degree and magnitude of differential selection acting upon populations across a species’ range (Guderyahn et al. 2023). In addition, turtles of different age classes vary in their habitat use as well as area used, both important considerations for wildlife conservation and habitat management (Bernstein et al. 2023).

In 1993, one of us (R.R.R., unpubl. data, 1993) began marking and measuring ornate box turtles in eastern Iowa. R.R.R. took high-quality photographs of turtles from multiple angles around their first capture, which enabled later investigations into growth ring counts as a proxy for age. In this paper, we leveraged these historical photos to report minimum ages for ornate box turtles originally encountered by R.R.R. and recaptured in recent years.

Methods. — Ornate box turtles were monitored at 2 sites within the Hawkeye Wildlife Area, Johnson County, Iowa, USA, from 1993 to 2022. During this period there were 3 research teams, led by R.R.R. (1993–1996), N.P.B., S.A.M., and R.W.B. (1996–2018), and D.F.H. (2020–present). Our sites consisted of sandy soils that accumulated south of the Iowa River during postglacial flow and contain a mosaic of prairie, shrubs, wetland, hardwood forest, and surrounding corn and soybean agriculture (Richtsmeier et al. 2008). The core sand prairie area used by turtles at Mallard Pools was ∼ 2.75 ha of sand prairie, 24 ha of low-lying forest, and 0.65 ha of wetland; the sand prairie at Greencastle covered ∼ 25 ha, with the remainder in successional shrubs and hardwood trees.

We captured turtles during visual-encounter surveys nearly every year from 1993 to the present and with drift fences made of window screen (∼ 1.5 mm) mesh and 19-L bucket pitfall traps surrounding the main overwintering area from 2011 to 2018; we focused searches largely at Mallard Pools during 2011–2018. We checked buckets daily from mid-May to early July, at which time we removed the fences and buckets. Turtle-finding dogs were also employed for 4 consecutive days at both Greencastle and Mallard Pools in 2019, 2021, and 2022 (Boers et al. 2017); D.F.H. resumed visual-encounter surveys throughout the area in 2020–2022.

For each initial turtle capture, we notched marginal carapace scutes with a hacksaw blade using a coding system adapted from Cagle (1939) and determined sex based on a combination of characteristics including coloration, shell shape, cloacal location, tail length, and hind foot inner claw thickness (Legler 1960). Carapace and plastron measurements were made with Vernier and digital calipers (± 0.1 mm).

We estimated age in 2 ways. First, for turtles recaptured between 2021 and 2022 that were originally marked and photographed (by R.R.R.), 4 of us (N.P.B., S.A.M., T.J.V., and D.F.H.) independently counted plastral rings from photographs of turtles taken near the time of their first capture (1996–1998); these counts were used for our analysis. We took a conservative approach to our estimates and used the lowest counts if growth lines were deemed ambiguous. We then averaged the 4 independent plastral scute count estimates and rounded up to the nearest integer to estimate that turtle’s age at initial capture, which we related to the year of their recapture. We concentrated our age estimates on these most recently recaptured individuals.

Second, we also estimated how many additional turtles survived into their 30s and 40s that were recaptured between 2000 and 2020 using plastral photographs from which rings could be counted, presence of a worn plastron at first capture, and/or initial carapace and plastron measurements as approximate indicators of age.

Results. — Between August 1993 and September 1996, 374 turtles were marked and photographed by R.R.R. (6 unknown sex, 178 males, and 190 females). We recaptured 30 turtles between 2021 and 2022 that were originally marked in 1993–1996 (Table 1). Assuming plastral rings approximate age, we estimated that 13 of these turtles were in their late 30s at the time of recapture and 17 were in their early 40s (Table 1). Two females (11, 192) clearly grew between first capture and 2021 and 2022, respectively, but initial and final measurements indicate that the remainder of turtles were adults (at least 15–16 yrs old) when first marked (Table 1).

Table 1. Minimum age estimates of 30 ornate box turtles (Terrapene ornata) from an Iowa population first marked between 1993 and 1996 and recaptured between 2021 and 2022. Straight-line-carapace (SCL), and plastron (PL) measurements (mm) from time of original (0) and recapture (R).

View Fullscreen

Of the originally marked turtles, 57 turtles were not recaptured after 1999 and not included in the analysis. Of the remaining 317 turtles, 122 females were recaptured (90 between 2000 and 2010 and 78 between 2011 and 2022, including some of the same individuals from the previous time period). For males, 132 turtles were recaptured (103 between 2000 and 2010 and 93 between 2011 and 2022). Forty-eight (13%) of the originally marked turtles were never recaptured, 15 were found dead, and a poacher removed an unknown number of turtles from the population in 2001. Therefore, in addition to the turtles listed in Table 1, we estimate that another 2 females survived into their 40s and another 61 males and 72 females survived into their 30s. Therefore, using both methods of age estimation, of the 374 turtles marked between 1993 and 1996 that were recaptured between 2000 and 2022, at least 165 (44%) lived into their 30s and, not exclusively, at least 13 (3.5%) lived into their 40s.

Discussion. — Although survivorship of adult box turtles is high (e.g., Redder et al. 2006; Currylow et al. 2011), few individuals actually live to old age (Gibbons and Semlitsch 1982). In agreement with Lewis and Iverson (2018), we found that previous conclusions of ornate box turtles rarely living past 30 yrs in the wild do not apply to our data. We stress that our estimates should be taken as potential age in years because of controversy regarding whether growth rings are accurate indicators of age (Wilson et al. 2003).

From a sample size of 609 individuals, Lewis and Iverson (2018) estimated at least 5 survived to 33 yrs, 4 to 36 yrs, and 11 lived to 45 to 57 yrs. They used counts of plastral growth rings for their estimated ages at initial capture, but all of their initial ages for the 11 turtles with longevity records were listed as “> 20” yrs (Lewis and Iverson 2018; their table 1), which suggests that the shells were smooth (or “worn”) because precise numbers were not indicated. In contrast, we leveraged historical photos taken around first capture, and, based on Bernstein et al. (2018), we conservatively assigned an age of 16 yrs for adult turtles from which we could not get accurate growth ring counts. Therefore, our estimates are likely minimum ages (i.e., 40 +) because some original photographs of plastrons showed wear indicative of much older turtles with few or no growth rings visible (“Worn” in Table 1). Males in Kansas were reproductive at plastral lengths of 100–109 mm (8–9 yrs) and females matured at plastral lengths of 110–119 mm (10–11 yrs; Legler 1960). Therefore, carapace and plastral measurements indicated most turtles were reproductive adults at first capture (Table 1; Bernstein et al. 2018). Based on total sample size of Lewis and Iverson (2018), our percentage survival into the 30s is larger, but our data agree that females may live longer than males in wild populations, or at least, survive longer on average. Continued study of this Iowa population may definitively reveal some turtles living to 50 yrs.

In Illinois, Edmonds (2020) described a similar long-term mark–recapture study and suggested that the oldest ornate box turtle was at least 31 yrs old, originally marked in 1988 as a presumed adult and recaptured in 2019. However, without age estimates at first capture, only speculation could be made that turtles lived over 40 yrs (Edmonds 2020). Our study provides additional evidence of lifespans over 40 yrs for maximum longevity in wild ornate box turtles.

That few ornate box turtles survive to old age is not unexpected (Graham and Hutchison 1969; Gibbons 1987; da Silva et al. 2022), but knowledge of lifespan and population demography are important for understanding conservation genetics of the species, interspecies ecological comparisons, population viability, land management, and the evolution of aging (Kuo and Janzen 2004; Reinke et al. 2022). The most immediate conservation implication of accurate knowledge of maximum lifespan is for population viability analyses, which is a modelling tool that uses life-history information as input data to estimate the future size and risk of extinction for populations (Boyce 1992). The ornate box turtle is threatened in many of the states where it occurs (Redder et al. 2006); and thus equipped with new knowledge about its longevity, we can provide more accurate estimations of population vital rates and long-term viability (e.g., King et al. 2021).

Another obvious implication of increased lifespan is the reproductive potential of older individuals. Although turtles continue to reproduce as they age, Warner et al. (2016) found older painted turtles (Chrysemys picta) had decreased reproductive success. As one of only two large populations of ornate box turtles in Iowa, our previous studies focused primarily on threats of nest predation and habitat loss. Now, the presence of older individuals must also be considered in developing multifaceted conservation plans. As noted by Tinkle (1979), the benefits of long-term research projects are invaluable as we continue to ascertain the local ecology and conservation needs of threatened species.