Chersina angulata has an unusual reproductive pattern for a tortoise inhabiting climatic zones ranging from winter rainfall with extreme aridity in the northwest to temperate with all-year rainfall in southern South Africa. Females are reproductively active over a prolonged breeding season during which they ovulate sequentially and oviposit multiple clutches, usually containing a single egg, occasionally two. Ovulations can occur from late January to late November and eggs are deposited from March to December. The shelling process of eggs in the oviduct takes place in all seasons, including winter, with only a short hiatus in mid-summer (Hofmeyr 2004, 2009).

A generally accepted wisdom for the order Testudines is that embryonic development in the oviducts is arrested at the gastrula stage (Ewert 1985; Andrews and Mathies 2000; Rafferty and Reina 2012). However, C. angulata can retain eggs with developing embryos until hatching normally occurs in autumn (Hofmeyr and Kuchling 2017; Kuchling and Hofmeyr 2022) and preovipositional developmental arrest is facultative (Kuchling and Hofmeyr 2023). Early developmental arrest in chelonians may be due to insufficient oxygen to support embryonic development in the uterus (Rafferty et al. 2013; Blackburn and Sidor 2014), and thickness and structure of the eggshell is one factor that may influence oxygen supply to the embryo (Andrews and Mathies 2000). Egg shell properties appear to be critical to allow embryo development in retained eggs, but eggshell characteristics of C. angulata are unknown. In this article we test the hypothesis that egg shells of C. angulata may be relatively thin for the family Testudinidea. We analyze data on oviducal eggshell calcification/thickness and egg retention time prior to oviposition in relation to season and temperature collected during an earlier reproductive study of captive C. angulata, but not reported in Hofmeyr (2004).

Methods. — We used ultrasound scanning to study follicle and egg production in wild C. angulata females at Dassen Island (33°26′S, l8°05′E) and the West Coast National Park (33°13′S, l8°09′E) in the southwestern Cape, South Africa, over 9 seasons from March 1999 to April/May 2001. To allow more-regular ultrasound scanning, and a comparison of ultrasound with radiography results, 16 females and 5 males were translocated in August 1999 from the WCNP to the home of M.D.H. at Kuils River (33°56′S, l8°41′E). The captive tortoises were kept in an outdoor enclosure (10 × 20 m) containing shrubs and lawn grass for shelter and food and an area of bare soil for easy nesting. The tortoises receive supplemental food (vegetables and greens) 2–4 times per week, depending on season.

From August 1999 to December 2001 (29 mo), M.D.H. used a Phillips PIE Veterinary ultrasound scanner (5 and 7.5 MHz frequency) with a curvilinear, intravaginal probe (14.4 × 21.6 mm) to examine the captive females at least twice a month. Females were scanned through the cranial and left and right inguinal openings, while held in a water bath (Henen and Hofmeyr 2003). The number and size of follicles were recorded with each assessment and when ovulation became imminent, the female was scanned daily or every second day to record the time required from ovulation for albumen secretion, the formation of the eggshell membranes (thin to thick), and membrane calcification as indicated by increasing cloudiness of the albumen and yolk. To verify the number, position, and calcification stage of eggs, females were radiographed (200 mA and 50 kV for 0.16 sec at a distance of 90 cm) once a month during the first year of study. We distinguished 5 radiographic stages of eggshell calcification: 1) barely visible when using a light table, 2) showing a faint outline with the naked eye, 3) clearly visible with the naked eye but with a thin shell, 4) more prominently calcified, and 5) over-calcified.

Results. — Following ovulation, albumen secretion and membrane formation in the oviduct were relatively quick (Table 1), and the time needed for the first 4 stages evaluated by ultrasound scanning did not differ between cold (May–September) and warm (February–April and October–November) months (t- and Mann-Whitney tests: p > 0.52). The time required for the onset of eggshell calcification (ca. 9 days) corresponded for ultrasonography (albumen cloudy) and radiography (radiograph stage 1) results (Table 1).

Table 1. Average number of days from ovulation to specific stages in egg development of captive Chersina angulata as determined by ultrasound scanning (first 4 stages) and radiography (last 5 stages). See text for a description of development stages. Note that the 16 females were radiographed monthly from August 1999 to August 2000 and thus a specific egg often appeared in more than 1 successive radiograph. SD = standard deviation.

View Fullscreen

When analyzing radiography data, we excluded stages 1 and 5 because low sample size did not allow tests for interactions between 2-monthly periods and radiography stages (Table 2). Two-way ANOVA results showed that retention time did not differ among periods (F_3,79 = 2.68, p = 0.053), was significantly different among stages (F_2,79 = 39.9, p < 0.001), and that there was an interaction between period and stage (F_6,79 = 2.24, p = 0.047). More time was required to achieve each successive radiography stage (4 > 3 > 2) for results overall as well as results within the last 2 periods of the year (July–August and September–October). However, within the first 2 periods of the year (March–April and May–June), the time needed to reach calcification stages 3 and 4 did not differ (Table 2).

Table 2. The frequency (%) of radiography stages 2–4 and associated mean ± SD retention time of Chersina angulata eggs at different periods of the year. Mean temperatures in the 4 successive periods were 18.6°C, 14.5°C, 13.8°C, and 15.4°C.

View Fullscreen

Eggs ovulated in different months of the years differ in time to oviposition (1-way ANOVA: H₉ = 61.6, p < 0.001). Egg retention of October and November ovulations was shorter than eggs ovulated in February, April, May, and July. In most instances (71%), females laid their eggs when the eggshell still appeared relatively thin on the radiographs (stage 3), with fewer eggs becoming more prominently calcified (stage 4) before oviposition. The egg of only 1 female progressed to stage 5. There was no difference among periods (time of year) for eggs that progressed to stage 3 or 4 before oviposition (1-way ANOVA, p = 0.21).

Discussion. — Chersina angulata egg retention times differed among sequentially ovulated eggs from February to November, being generally shorter in eggs ovulated in late spring than in eggs ovulated in late summer, autumn, and winter. On the other hand, more time was required to achieve eggshell calcification from mid-winter to spring (July–October) than in autumn and early winter (March–June). In most instances (71%), females laid their eggs when the eggshell still appeared relatively thin on the radiographs, with only few eggs becoming more prominently calcified prior to oviposition. The egg shells of C. angulata are generally described as brittle, but rather thin for a tortoise, and some females frequently break their freshly laid egg when closing the nest hole (Pfau 2022).

Regarding in uterus embryonic development beyond the gastrula stage in C. angulata (Hofmeyr and Kuchling 2017; Kuchling and Hofmeyr 2022, 2023), squamates that lay advanced embryos generally have thinner eggshells than turtles (Andrews and Mathies 2000). However, common wisdom suggests that, in the order Testudines, the thinnest and least-calcified eggshells are produced by species that nest in humid substrates and have short incubation periods, whereas species from arid and semiarid regions with long incubation periods produce eggs with thick, calcareous shells (Miller and Dinkelacker 2008). Given this assumption, the arid-adapted C. angulata would be an unlikely candidate for advanced embryonic development in eggs retained inside the female. However, in 2 species of Chersina's arid-adapted sister genus Chersobius, C. signatus and C. boulengeri, eggshells are thinner than recorded in any other tortoise genus and eggshells have a relatively simple structure, without multiple layers of crystallites or an outer cuticle layer (Loehr et al. 2019). As opposed to Miller and Dinkelacker's (2008) assumption, a recent review of eggshell structure evolution in reptiles is in agreement with these results by showing that, as a general pattern for reptiles including turtles, species nesting in environments with higher precipitation produce eggs with thicker eggshells than those in arid regions and that eggshell thickness does not increase with ground temperature or incubation length (D'Alba et al. 2021).

Our radiographic results indicate that C. angulata egg shell thickness is also in agreement with the findings of D'Alba et al. (2021) for species inhabiting arid and semiarid regions: the shells of the majority of eggs appear relatively thin at oviposition (Table 2). Chersobius signatus is sympatric with C. angulata along the Cape west coast and the distribution of Chersobius boulengeri partly overlaps with that of C. angulata. Although eggshells of C. angulata have not yet been studied in detail, they may have attributes comparable to those of C. signatus and C. boulengeri. Such egg shell characteristics may improve oxygen diffusion to developing embryos in utero. Features of the oviduct wall and oviduct secretions should also be important for gas exchange, but have not yet been investigated.

A limitation of the present study is that it did not include eggs ovulated during November, the most likely candidates for the phenotypically plastic abolishment of preovipositional developmental arrest (Kuchling and Hofmeyr 2023) and retainment inside the female until the completion of embryonic development in autumn (Hofmeyr and Kuchling 2017; Kuchling and Hofmeyr 2022). It only includes eggs ovulated from February to October that incubated in nests. Nevertheless, it is interesting that, with 1 exception, eggs did not become over-calcified, despite being retained for extended periods. In most instances (71%), females laid their eggs when the eggshell still appeared relatively thin on the radiographs (stage 3), with fewer eggs becoming more prominently calcified (stage 4) before oviposition. Thinner eggshells would hold an advantage if females retain eggs in utero to full term. Overall, the ability to retain a relatively thin eggshell over a long time may be one of the traits that allow the evolution of facultative viviparity in C. angulata (Hofmeyr and Kuchling 2017; Kuchling and Hofmeyr 2022, 2023).