The Rio Grande cooter (Pseudemys gorzugi) is an imperiled freshwater chelonian native to the southwestern United States and northeastern Mexico (Ernst and Lovich 2009). Pseudemys gorzugi is the westernmost representative of the genus Pseudemys and has a relatively small distribution. Within the United States, P. gorzugi occupies the Pecos, Black, and Delaware rivers in New Mexico (Degenhardt et al. 1996) and the Pecos, Rio Grande, and Devils rivers in Texas (Dixon 2013; Bailey et al. 2014). Despite an apparent gap in the distribution of P. gorzugi in the lower Pecos River of west Texas (Ernst and Lovich 2009; Dixon 2013; Mahan 2022; Mahan et al. 2022), the species has minimal genetic structure across this portion of its distribution (Bailey et al. 2008, 2014), suggesting only recent allopatry. Permanent, spring-fed Texas creeks that drain into the aforementioned river systems, such as Independence, San Felipe, and Las Moras creeks, also are known to harbor P. gorzugi (Dixon 2013; Pierce et al. 2016). Within Mexico, populations are known from the states of Coahuila, Nuevo León, and Tamaulipas (Pierce et al. 2016). Currently, P. gorzugi is listed as threatened in both New Mexico and Mexico (Secretaría de Medio Ambiente y Recursos Naturales 2010; New Mexico Department of Game and Fish 2018) and is a species of greatest conservation need in Texas. Threats to the taxon include habitat degradation (Pierce et al. 2016), commercial collection (Bailey et al. 2008; Dixon 2013), roadway mortality (Walls 1996), recreational fishing (Waldon et al. 2017; Bassett et al. 2020), and wanton shooting (Christ-man and Kamees 2007; Suriyamongkol et al. 2019). Although once considered poorly studied (Ernst and Lovich 2009), P. gorzugi has received increased research attention in the past decade, with multiple works addressing various aspects of the taxon's distribution and natural history (e.g., Brush et al. 2017; Letter et al. 2019; Bassett 2021; Sirsi 2021).

A relatively large proportion of turtle species (52%) are threatened (Rhodin et al. 2018), primarily because chelonian populations are sensitive to increases in adult mortality. This sensitivity is a consequence of delayed sexual maturity and low annual recruitment (Brooks et al. 1991; Congdon et al. 1993, 1994). Information on nesting phenology, reproductive output, and size at sexual maturity is therefore pertinent to species conservation and management; however, knowledge gaps exist for many chelonian species, including P. gorzugi. Work to describe the reproductive biology of P. gorzugi has been conducted previously. At the Black River, in May of an unspecified year, Degenhardt et al. (1996) captured a single gravid female containing 9 oviductal eggs that averaged 42 mm long and 31 mm wide. Lovich et al. (2016) later captured a gravid female P. gorzugi in June 2012 that contained 10 eggs, followed by Letter et al. (2017) who captured a gravid female in June 2017 containing 12 eggs that averaged 40.3 mm long and 31.1 mm wide. Suriyamongkol and Mali (2019) conducted an investigation into the reproductive biology of P. gorzugi during which 86% of gravid females were observed within a 6-wk period spanning May and June. A mean clutch size of 9.3 eggs (range, 5–14 eggs) was observed for this population. Female plastron length (PL) had a positive linear correlation with both clutch size and mean egg width. Finally, double clutching occurred in 19% of gravid females (n = 16) in their study. These efforts have focused on the Black River population in southeastern New Mexico and the results for the species are expected to vary in other watersheds.

The reproductive characteristics of chelonian taxa are known to vary across geographic gradients. Such characteristics include clutch size and frequency (Tinkle 1961; Gibbons 1983; Litzgus and Mousseau 2006; Ashton et al. 2007), size at sexual maturity (Gibbons et al. 1981; Litzgus and Brooks 1998), and nesting phenology (Iverson et al. 1997). Hypothesized drivers of differential reproductive biology include climatic variables, particularly rainfall and temperature, and site-specific differences in resource availability (Ernst and Lovich 2009). Considering the propensity for the reproductive biology of chelonian taxa to vary across geographic space, and that the distribution of P. gorzugi covers at least 3 unique climatic zones (Fovell and Fovell 1993; Peel et al. 2007), further study of life history characteristics of this turtle species is needed. Climatic zones in Texas as referenced are defined according to temperature and precipitation differences. Our goals were to supplement the existing body of knowledge by sampling P. gorzugi across 3 sites in western Texas and recording clutch size, clutch frequency, length of breeding season, size of females at sexual maturity, and phenology of egg development.

METHODS

Field Sites and Sampling Regimen. — We sampled P. gorzugi at 3 sites in western Texas (Fig. 1) from 31 March 2017 to 12 August 2019: the Devil's River at Dolan Falls and Finnegan Springs, San Felipe Creek in Del Rio, and Las Moras Springs in Brackettville. These sites are located at the interface of the Edward's Plateau and Chihuahuan Desert ecoregions (Blair 1950) and are characterized by high water clarity. All sites represent riverine habitats generated by spring discharge from the Edwards Aquifer and include a combination of lotic and lentic microhabitats. Dolan Falls and Finnegan Springs are separated by a distance of ca. 2.9 river km. Based on the documented long-distance movement of P. gorzugi (MacLaren et al. 2017b; Sirsi 2021), we combined all data collected at these 2 sites into a singular data set (hereafter, Dolan Falls). Ultimately, the sampling site at Dolan Falls was ca. 3.8 river km long from Finnegan Springs to ca. 0.9 river km downstream of the waterfall at Dolan Falls. We sampled San Felipe Creek from the San Felipe Springs Golf Course to the train crossing immediately south of US Highway 90, a total distance of ca. 0.38 river km. We sampled Las Moras Springs from the headwaters that form a deep pond to a small artificial impoundment downstream, a total distance of ca. 5.5 river km.

View Figure

• Download as Powerpoint
• Download Figure

During the summer of 2017, we opportunistically sampled San Felipe Creek in Del Rio for gravid females. During the summer of 2018 (April–September), we sampled each site at least once a month and on 3 occasions we sampled some of the sites twice (San Felipe Creek was sampled twice in April and both San Felipe Creek and Las Moras Springs were sampled twice in July). After September 2018, we continued to sample Las Moras Springs on a monthly basis until March 2019. From March through August 2019, we sampled Las Moras Springs biweekly.

Turtle Capture and Processing. — We captured turtles by hand and with basking traps and hoop net traps baited with sardines. Hoop net traps were 76 cm in diameter, fiberglass, with a single opening and wide mouth with 2.5-cm mesh size and 4 hoops per net (Memphis Net and Twine Company, Memphis, TN). Basking traps were constructed of 1.5-cm plastic mesh and 10.16-cm-diameter polyvinyl chloride pipe and measured 100 cm long, 72 cm wide, and 42 cm deep. We used traps from October 2018 to February 2019 and we conducted hand capture throughout the duration of the study. We shell-notched captured turtles in cohorts based on the year of capture (Cagle 1939) and tagged turtles using subcutaneous passive integrated transponder tags (Avid Identification Systems, Inc., Norco, CA). We weighed turtles using a digital scale and measured straight-line shell dimensions using Haglöf tree calipers including midline carapace length (CL; Iverson and Lewis 2018), maximum carapace width (CW; van Dam and Diez 1998), midline PL (Iverson and Lewis 2018), maximum plastron width (PW; measured as distance from the lateral margin of each abdominal scute), and maximum body depth (BD; van Dam and Diez 1998). Males and females were both targeted for capture and processed at San Felipe Creek and Dolan Falls, but due to personnel and time constraints, females were prioritized at Las Moras Springs (i.e., males were oftentimes ignored while snorkeling or, if captured, not subsequently processed).

We used an Ibex Pro ultrasound machine (E.I. Medical Imaging, Loveland, CO) at the capture site on every female turtle captured that was large enough for the probe to be properly inserted into the inguinal pocket (inguinal pocket larger than ca. 2 × 3 cm). To maximize depth and clarity of sonogram imaging, we used a gain of 27 dB, a near setting of 40 dB, and a far setting of 46 dB. During the ultrasound, we measured maximum follicle diameter. Although ultrasonographic imagery has previously been used to estimate clutch size, such estimates were done when imagery was available from both the left and right inguinal regions of a turtle (Zuffi et al. 2015). Because we only took a single ultrasonographic image per turtle, it was not possible to accurately obtain a total count via sonogram imaging.

We transported gravid turtles to Val Verde Veterinary Hospital in Del Rio, Texas, to subject them to radiography. We radiographed turtles from only San Felipe Creek and Las Moras Springs due to the close proximity to the veterinary facility. Eggs could be accurately differentiated from follicles during the ultrasound because eggs were larger with a clearly visible calcified membrane around the yolk (Fig. 2). We positioned turtles carapace side down for radiograph imaging so that an image was taken through the plastron. We found that shelled eggs were most visible at 70 kVp and 100 mA (Jenkins 2006).

View Figure

• Download as Powerpoint
• Download Figure

Statistical Analysis. — We conducted all statistical analyses in Program R, version 4.1 (R Core Team 2021). We conducted a linear regression on data pooled from San Felipe Creek and Las Moras Springs to determine whether female PL was correlated with clutch size. In one instance, we obtained 2 clutch size measurements for a single turtle (ID 467A510A00) across a 29-d period. We think these represent unique clutches (see “Discussion”) and therefore decided to include both data points in the regression analysis. We used PL as the predictor variable and clutch size as the dependent variable. We log₁₀-transformed both variables to increase normality in the data set and allow comparability with other studies (King 2000; Ryan and Lindeman 2007; Suriyamongkol and Mali 2019). We conducted a linear regression to determine whether female PL was significantly correlated with mean egg width. We measured egg width using ImageJ v. 1.53k. We set PL as the predictor variable and mean egg width as the dependent variable. We log₁₀-transformed both variables. We conducted a quadratic regression to evaluate the correlation between maximum follicle diameter and day of year using data pooled from all sites. Because we observed a quadratic trend with follicles reaching their smallest size in late October, we set 27 October as 0 and 26 October as 364. We used day of year (i.e., days elapsed since 27 October) as the predictor variable and maximum follicle diameter as the dependent variable. We conducted a binomial logistic regression with a quadratic term to predict egg presence in female P. gorzugi with ordinal date (i.e., days elapsed since 31 December) set as the predictor variable. If eggs were detected in any female on a given ordinal date, we assigned the response variable a value of 1. If eggs were not detected in any females ≥ 179 mm PL (minimum known size for sexual maturity in female P. gorzugi, see “Discussion”) on a given ordinal date, we assigned the response variable a value of 0. We assessed model fit using Nagelkerke's r².

RESULTS

San Felipe Creek. — At San Felipe Creek, we captured 36 individual P. gorzugi, including 17 males and 19 females (1.12 females for every male). Our total number of captures (including recaptures) was 54. Of the females captured, 2 were captured twice, 1 was captured 3 times, and 1 was captured 5 times. We detected follicles in 8 females. We detected 7 gravid females with shelled eggs and obtained reliable radiographs from 3 of those turtles. Radiographs from the other 4 turtles showed no presence of eggs, indicating that although the eggs were mature in size and detectable by the ultrasound, sufficient calcification had not yet taken place for detection via radiograph imaging. The 3 females that did show eggs via radiograph imaging revealed clutch sizes of 8, 10, and 12 eggs. This resulted in a mean ± 1 SD clutch size of 10.0 ± 2.00 SD. Mean egg width was 31.33 ± 2.11 mm SD and maximum follicle diameter ranged from 12 to 30 mm. The earliest and latest detections of shelled eggs were on 7 April and 3 August, respectively. The smallest female gravid with shelled eggs had a PL of 179 mm.

Las Moras Springs. — At Las Moras Springs, we captured 59 individual P. gorzugi, including 16 males, 42 females, and 1 juvenile (2.63 females for every male). Our total number of captures (including recaptures) was 118. Of the females captured, 3 were captured twice, 4 were captured 3 times, 3 were captured 4 times, 2 were captured 5 times, 4 were captured 6 times, and 1 was captured 12 times. We detected follicles in 35 females. We detected 11 females gravid with shelled eggs and obtained reliable radiographs from 7 of those turtles, which included clutch sizes of 8, 12, 13, 13, 15, and 17. One female (ID 467A510A00) was captured on 7 June 2018 and then recaptured on 6 July 2018. On 7 June 2018, her radiograph indicated a clutch size of 7, whereas on 6 July 2018, a second radiograph indicated a clutch size of 10. Presuming that the 2 radiographs represent unique clutches, the mean clutch size for this population is 11.9 ± 3.40 SD. Mean egg width was 30.96 ± 1.71 mm SD and maximum follicle diameter ranged from 7 to 26 mm. The earliest and latest detections of shelled eggs were on 7 April and 12 August, respectively. The frequency of females gravid with shelled eggs was highest in the months of May 2018, July 2019, and August 2019 (Fig. 3). The smallest female gravid with shelled eggs had a PL of 241 mm.

View Figure

• Download as Powerpoint
• Download Figure

Dolan Falls. — At Dolan Falls, we captured 245 individual P. gorzugi, including 194 males, 47 females, and 4 juveniles (0.24 females for every male). Our total number of captures (including recaptures) was 334. Of the females captured, 11 were captured twice and 1 was captured 3 times. We detected follicles in 26 females. We detected 3 females gravid with shelled eggs. Egg length ranged from 31 to 42 mm and maximum follicle diameter ranged from 9 to 27 mm. The earliest and latest detections of shelled eggs were on 21 April and 5 June, respectively. The smallest female gravid with shelled eggs had a PL of 242 mm.

All Sites. — Clutch size ranged from 7 to 17 with a mean of 11.4 ± 3.11 SD; n = 11). The smallest female we captured that was gravid with shelled eggs measured 179 mm PL. The mean PL of females gravid with shelled eggs was 247.3 ± 20.17 mm SD (n = 21). Mean egg width was 31.05 ± 1.81 mm SD (n = 126) and ranged from 26.77 to 35.48 mm. The regression between PL and clutch size in female P. gorzugi was nonsignificant (r² = 0.20, F_1,9 = 2.31; p = 0.16). The quadratic model evaluating the correlation between day of year and maximum follicle diameter was significant (r² = 0.22, F_2,100 = 15.61; p < 0.01; Fig. 4). The binomial logistic regression evaluating the correlation between ordinal date and egg presence (Fig. 5) showed that the effect of ordinal date was nonsignificant (p > 0.05; Table 1) and that the quadratic effect of ordinal date was significant (p < 0.05; Table 1). Probability of detecting a female gravid with shelled eggs was greatest on ordinal day 161, 10 June (probability = 67.01%; Fig. 5). Nagelkerke's r² for this regression was 0.37. There was no relationship between PL and mean egg width per clutch (r² = 0.006, F_1,9 = 0.052; p = 0.83).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Table 1. Results of the binomial logistic regression that evaluated the correlation between ordinal date and egg presence in female Pseudemys gorzugi captured in west Texas from 2017 to 2019, including model covariates, estimated coefficients, standard errors, and p-values. Ordinal date had a nonsignificant effect and ordinal date squared had a significant effect (p < 0.05). Nagelkerke's r² for this regression was 0.37.

View Fullscreen

DISCUSSION

Our results demonstrate that P. gorzugi in western Texas carry eggs from early April until mid-August (Fig. 3). Previous to our study, the earliest and latest known detections for P. gorzugi gravid with shelled eggs were 15 May and 21 June, respectively (Suriyamongkol and Mali 2019). Our findings indicate that the nesting season for P. gorzugi is longer than previously documented, at least for areas in the southcentral portion of its distribution. Although the longer nesting season observed here may be due to geographic variation in the reproductive biology of P. gorzugi, it is more likely an artifact of our more temporally extensive sampling. For example, Lovich et al. (2016) limited sampling to June. Similarly, Suriyamongkol and Mali (2019) limited sampling to the summer months of May, June, July, and August, along with a 5-d bout of sampling in mid-March. Suriyamongkol and Mali (2019) observed 86% of all the gravid females they captured within a 6-wk period spanning May and June. Similarly, our binomial logistic regression predicted the highest probability (67.01%) of capturing a sexually mature female gravid with shelled eggs occurs on 10 June (Fig. 5), further indicating May–June as the height of P. gorzugi egg production. Obtaining a clear understanding of the nesting season length throughout the range of P. gorzugi will require additional studies with year-round sampling effort.

We were able to collect reliable clutch-size data from 10 turtles. We did not observe a relationship between PL and clutch size as reported in the Black River population by Suriyamongkol and Mali (2019). This finding runs contrary to what is typically observed with emydid turtles, where a strong relationship between the 2 variables often exists (Jackson 1988; Dodd 1997; Zuffi et al. 1999). One of the turtles we radiographed (ID 054895057) was carrying 17 shelled eggs, the maximum reported clutch size record for this species (Fig. 2). The mean clutch size observed in our study was 11.4 ± 3.11 SD, which is greater than the mean clutch size of 9.3 ± 2.89 SD observed by Suriyamongkol and Mali (2019) in the Black River. It is possible that mean clutch size for P. gorzugi varies across geographic space, a phenomenon that has been observed in other chelonian taxa such as Sternotherus odoratus (Tinkle 1961), Chrysemys picta (Rowe 1994), Chelydra serpentina (Iverson et al. 1997), Clemmys guttata (Litzgus and Mousseau 2006), Gopherus polyphemus (Ashton et al. 2007), and Malaclemys terrapin (Lovich et al. 2019). Although it has been argued that turtle clutch sizes may decrease with latitude (Ashton et al. 2007), we do not observe that pattern when we compare our findings to those of Suriyamongkol and Mali (2019). However, across these 2 studies, only 26 clutch size measurements exist (15 from New Mexico, 11 from Texas). Additionally, we found no relationship or directionality between egg width and turtle size, in contrast with findings from New Mexico (Suriyamongkol and Mali 2019). Failure to realize these 2 patterns may be an artifact of limited sample size (n = 11) in our study, or the relatively small difference in latitude between these 2 populations (approximately 2°), compared with more geographically widespread species such as Trachemys scripta. For example, Ashton et al. (2007) found clutch-size variation for G. polyphemus across approximately 5° of latitude. Perhaps if P. gorzugi reproductive biology was studied across the entirety of its distribution, a clear relationship between latitude and clutch size would be observed. Alternatively, the lack of a relationship could suggest that there may be an optimum egg size that each clutch approximates regardless of female dimensions (Ryan and Lindeman 2007). It is important to note that we were able to measure egg width from radiographs, rather than directly measuring egg mass by weight. Egg width may not be a reliable proxy for egg mass, or in other words, larger females might lay larger eggs. This pattern is simply not detectable using the methods available to us during this study.

By naïve comparison to Suriyamongkol and Mali (2019), P. gorzugi in Texas appear to have a greater reproductive output than turtles in New Mexico. Average clutch size in Texas exceeded New Mexico by approximately 2 eggs while average egg width in Texas was approximately 1 mm larger. In addition, gravid turtles in Texas were on average larger (PL) than those in New Mexico. However, we cannot remark on whether these differences are statistically significant at this time. Future research should seek to quantify if these reproductive traits vary predictably across geographic space. Pseudemys gorzugi in Texas may be able to optimize egg size and actualize a greater reproductive output by attaining larger body size more rapidly. The last in turn may be due to underlying climatic differences (Fovell and Fovell 1993) or site-specific differences in the seasonal abundance of nutritional resources.

One female P. gorzugi (ID 467A510A00) from Las Moras Springs was captured on 7 June 2018 and then recaptured on 6 July 2018. On 7 June 2018, the radiograph indicated a clutch size of 7, whereas 29 d later, a second radiograph indicated a clutch size of 10. We think this case represents an instance of production of 2 clutches, given that internesting intervals for congeneric and confamilial chelonians are less than 29 d in length. For example, the average internesting interval for Pseudemys concinna suwanniensis is 21 d (Jackson and Walker 1997). Minimum internesting intervals for C. picta and Terrapene carolina are 9 and 22 d, respectively (Ewing 1935; Ream 1967). Additionally, in the Great Lakes region, C. picta have been observed with an internesting interval ranging 13–25 d, contingent upon ambient temperature (Edge et al. 2017). Suriyamongkol and Mali (2019) documented 3 instances of P. gorzugi producing 2 clutches in the Black River, making this the second published report of 2 annual clutches for the species.

Among our various study sites, we observed markedly different estimates of sex ratio. At San Felipe Creek, Las Moras Springs, and Dolan Falls, we observed sex ratios of 1.12, 2.63, and 0.24 females for every male, respectively. The skewed ratio at Las Moras Springs is easily explained because we released many captured males at this site without processing due to time and personnel constraints, while females captured at this site were always processed. However, the reason for the dramatically different sex ratios at San Felipe Creek and Dolan Falls are much less clear. Many studies point to female-skewed road mortality as a contributor to increased proportion of males among aquatic turtle species (Steen and Gibbs 2004; Aresco 2005). However, this hypothesis would suggest that males would be more common than females at San Felipe Creek than at Dolan Falls, since the former is within a city and the latter is quite remote. Our data instead show the opposite pattern. Anecdotally, we can confirm that female mortality at Dolan Falls likely does occur during nesting via predation, as indicated by the many females we have observed missing hind limbs. Additionally, it is important to note that Dolan Falls was sampled using hand capture exclusively, with a high degree of success relative to our other sites, where the turtles' strong response to divers in the water required the use of alternative trapping approaches. Many studies have shown that utilizing a single trapping technique can bias estimates of sex ratio among chelonians (e.g., Ream and Ream 1966; Tesche and Hodges 2015), and such bias should be examined further with respect to Texas populations of P. gorzugi. Mali et al. (2018) demonstrated that P. gorzugi in New Mexico are less likely to be caught by hand than trapped. An alternative hypothesis may be that temperature-dependent sex determination is responsible for the disparity we see among sites (Bull and Vogt 1979). Nesting habitat characteristics may vary among our sites, which could potentially lead to biased sex ratios at hatching. However, the geographic proximity of our sites would suggest that the thermal conditions of nesting habitat vary only marginally, making such a hypothesis unlikely.

There were a few anecdotal observations made throughout the course of this study regarding the reproductive biology of P. gorzugi. First, we observed male courtship behavior (i.e., trailing and foreclaw titillation) beginning as early as 12 January. Male courtship behavior was then continually observed from January into late August. Such prolonged courtship activity is also thought to occur in Pseudemys peninsularis (Ernst and Lovich 2009). We hand captured 2 hatchlings, one at Las Moras Springs and another along the Pecos River at Continental Ranch (29°51′45.4674″N, 101°24′0.288″W, datum WGS 84). The hatchling at Las Moras Springs (CL 34 mm, CW 34 mm, PL 33 mm, PW 25 mm, BD 18 mm, mass 8 g) was captured on 11 November 2018 during a visual survey. The marginal scutes were still pliable, suggesting recent emergence from a nest. The hatchling at Continental Ranch (CL 33.62 mm, CW 27.00 mm, PL 33.00 mm, PW 21.00 mm, BD 18.30 mm, mass 7.8 g) was captured on 9 July 2020 on a grassy ledge with sandy substrate ca. 20 m from the river margin. These incidental observations contribute knowledge where a paucity of data exist regarding the size of hatchlings and phenology of their occurrence in Texas populations of P. gorzugi (Bailey et al. 2014).

The smallest gravid female P. gorzugi we detected with shelled eggs measured 179 mm PL, which is 17 mm smaller than the smallest gravid female captured by Suriyamongkol and Mali (2019) in the Black River. Otherwise, females in our study that were gravid with shelled eggs were generally larger (> 30 mm on average). Size at sexual maturity may vary across a geographic gradient in this taxon, as seen in other chelonians such as C. guttata (Litzgus and Brooks 1998), M. terrapin (Seigel 1984), and T. scripta (Gibbons et al. 1981; Ernst and Lovich 2009). However, our results and those of Suriyamongkol and Mali (2019) are probably unreliable for inferring geographic variation for this attribute of P. gorzugi reproductive biology, given the limited sample size in both studies. Further study of P. gorzugi reproductive biology will be necessary to determine whether female size at sexual maturity truly varies across geographic space.

The ovarian cycle of chelonians can be divided into 4 stages: follicular enlargement, ovulation, oviposition, and quiescence. However, the timing of follicular development in turtles varies by taxon (Ernst and Lovich 2009). Our data suggest that follicle enlargement begins in the fall and continues through spring for P. gorzugi at our study sites (Fig. 4). As eggs are laid throughout the late spring and summer (Fig. 3), observed follicle size decreases, resulting in a quadratic trend at the population level (Fig. 4). Similar timing of follicle enlargement has been observed in Illinois and Louisiana populations of C. picta (Ganzhorn and Licht 1983).

Characterizing the reproductive biology of imperiled organisms is a vital step for informing conservation management. The necessity of doing so is especially true for chelonian taxa, which have low annual recruitment and delayed sexual maturity. The present investigation represents the first year-round study of P. gorzugi reproductive biology and delineates a possible nesting season of April–August for P. gorzugi in west Texas. This estimate provides management agencies an approximate timeframe during which protection measures can be implemented for gravid females, nest sites, and hatchlings, should such measures be necessary, until further research can facilitate producing finer-resolution life tables for this species. Furthermore, clutch size information can be utilized as a parameter in ecological models to estimate population growth or decline through time (Liu et al. 2002). Additional studies should be conducted in south Texas and Mexico to further elucidate any geographic variation that may exist in the reproductive biology of this imperiled turtle species. Methods for using global positioning system–enabled radiotelemetry units have been developed for P. gorzugi (MacLaren et al. 2017a). It would be meritorious to expand the use of telemetry to determine the terrestrial movement of gravid females, identify nesting sites, and describe the macro- and microhabitat characteristics of said nest sites for P. gorzugi (e.g., Walde et al. 2007; Tucker 2010; Macey 2015).