An unusually enlarged, broadened head, with expanded alveolar surfaces and hypertrophied jaw musculature (megacephaly), is not uncommon in turtles, being found in at least 19 genera across 9 families (Iverson et al. 1989). This morphology has been correlated with a durophagous diet (e.g., mollusks; Berry 1975; Patterson and Lindeman 2009). Megacephaly is universal in many taxa, suggesting a genetic basis, that is, a functional adaptation to a molluscivorous diet (Pfaller et al. 2011). However, in many taxa, megacephalic individuals are known in only some populations, and its expression is variable (e.g., Berry 1975; Dalrymple 1977; Iverson et al. 1989; Cann and Sadlier 2017), suggesting a developmental or environmental response to durophagy.

Megacephalic individuals of the Eastern Musk Turtle (Sternotherus odoratus) have been reported from North Carolina (no precise locality; Palmer and Braswell 1995); Brunswick County, Virginia (C. Hobsen, pers. comm., September 2017); and Comal County, Texas (Morrison et al. 2017). Given the rarity of these reports, it appears that megacephaly in S. odoratus may be an individual rather than a population-level (i.e., genetically based) phenomenon. For example, Morrison et al. (2017) reported the capture of both normal and megacephalic individuals at Comal Springs in Texas. Here, I report megacephaly in S. odoratus from central Arkansas as well as an attempted experiment to determine whether the condition is genetically based or induced by diet in this species.

In August 1987, I was informed by Bill Davis, a private reptile enthusiast from Hot Springs, Arkansas, that the Eastern Musk Turtles in the Glazypeau Creek drainage in Garland County, Arkansas, all had much-enlarged heads, unlike those in nearby drainages. Davis subsequently captured and shipped 24 Glazypeau Creek turtles to me for examination. These turtles were measured, preserved as vouchers, and deposited in the Carnegie Museum (CM 115906, 115908, 115949, 115951–54, 114956, 115907, 115950, and 115955) or the Florida Museum of Natural History (UF 71443–55). All were megacephalic (Figs. 1 and 2). Glazypeau Creek is a small, permanent, rocky, high-gradient stream that drains into Lake Hamilton, an artificial impoundment, at the southeastern margin of the Ouachita Mountains. The habitat in this stream is not typical for the species, which prefers soft-bottomed lentic systems with abundant vegetation (Iverson and Meshaka 2006; Ernst and Lovich 2009).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

To investigate the ontogeny of head size in S. odoratus, I attempted an experiment in which I raised hatchlings from central Arkansas from megacephalic females together with those from normal females. I captured 2 gravid female megacephalic musk turtles (maximum carapace length [CL] = 116 and 113 mm) by hand at night in a short tributary of Glazypeau Creek (3.7 km east of Highway 227 on Glazypeau Road; lat 34°35′29.8″N, long 93°07′48.7″W), east of Mountain Pine, Garland County, Arkansas, on 20 May 1997. On the same date, 2 gravid females with normal heads (CL = 111 and 108) were captured by hand at the Lonoke Fish Hatchery in a lowland, lentic fish pond (more typical habitat for the species) just south of Lonoke, Lonoke County, Arkansas (lat 34°46′06″N, long 91°54′40.5″W; ∼ 115 km away). All individuals originating from the Glazypeau Creek system (including the Lake Hamilton Fish Hatchery) were megacephalic, whereas such turtles were never encountered at the Lonoke locality (among several hundred observed). Measurements from both localities confirmed the much larger heads of Glazypeau turtles (Fig. 2). I have no data on diets in the 2 populations, but turtles from Glazypeau did defecate crushed pieces of mollusks after capture.

The gravid females from both sites were returned to Richmond, Indiana, and injected with oxytocin to obtain eggs (Ewert and Legler 1978). All eggs were measured, individually marked, placed in moistened vermiculite in the same plastic shoebox, and incubated together at 28 ± 1°C. Nine eggs from the 2 Glazypeau females averaged 29.54 mm in length (range, 28.1–31.2), 17.05 mm in width (16.7–17.5), and 5.41 g (5.14–5.75) in mass. However, the egg size data from the 2 Lonoke females were lost in a basement flood event in the Earlham Science Complex in the early 2000s, but 183 eggs from other females from Garland or Lonoke counties averaged 28.6 mm in length (24.7–32.6), 16.4 mm in width (13.7–18.5), and 4.88 g (2.63–6.59) in mass.

Of 9 eggs from the 2 clutches from megacephalic females, only 6 hatched, and of 8 eggs from normal females, only 4 hatched. On hatching, neonates were individually marked with a marginal notch and housed in individual plastic containers in an environmental chamber programmed on a 12:12 light:dark cycle with a daytime temperature of 25°C and a nighttime temperature of 20°C. Turtles were provided daily with large pellets of Purina catfish or trout chow (initially 1 pellet but increasing to 3–4 pellets over 2 decades), and the water was replaced with clean, deionized water weekly or biweekly, depending on water quality in the container. Hence, the diet consisted only of ostensibly soft food, and the turtles were never fed a hard food item (i.e., mollusk or crustacean).

My intent was to raise all of these hatchlings to adulthood under identical conditions, measuring maximum CL, maximum plastron length, and maximum head width approximately annually. Unfortunately, 2 environmental chamber failures over the course of the experiment compromised my ability to test for statistical significance in the difference between head width (relative to CL) between the offspring from megacephalic versus normal females. As a result, in June 2000, 3 Glazypeau and 1 Lonoke juvenile died, and in May 2007, 1 Glazypeau female and 2 Lonoke females died. The remaining 2 Glazypeau turtles (females) and 1 Lonoke female are still alive at the time of this writing.

Despite the small samples, all the surviving turtles exhibited the same ontogenetic change in head width, although their heads were wider than the parental non-megacephalic population and narrower than the megacephalic population (Fig. 3). This suggests that head width in this species is a plastic, environmental response to durophagy in at least Glazypeau Creek. The coexistence of both megacephalic and normal S. odoratus in Comal Springs in Texas (Morrison et al. 2017) provides additional support for this conclusion.

View Figure

• Download as Powerpoint
• Download Figure

In the only other study of head width versus body size in S. odoratus, Patterson and Lindeman (2009) demonstrated an increase in head width with body size across individuals in Pennsylvania that was correlated with an increase in mussels in the diet. However, head widths in that population never reached those in Glazypeau turtles (Fig. 2). Patterson and Lindeman (2009) suggested that the increase in head size in their population permitted the inclusion of more mollusks in the diet, but it may also be that head size in that population increased in response to increased mollusk consumption. Interestingly, Berry (1975) reported that in localities with low mollusk populations, the Loggerhead Musk Turtle (S. minor) had a reduced head width, suggesting an environmental effect on head width even in a closely related species presumed to have genetically based megacephaly.

The fact that captives from both Glazypeau and Lonoke developed heads of similar size but intermediate in size to their parents may suggest that the food pellets I provided may have been relatively harder than a typical diet but softer than a diet of mollusks. Future long-term experiments restricting diets of juvenile S. odoratus from megacephalic and non-megacephalic parents to mollusks versus soft-bodied prey (reviewed in Iverson and Meshaka 2006) could be even more informative.

Tribute to Peter C.H. Pritchard

During my first year of college in 1968, in a pet store in Omaha, Nebraska, I discovered the holy grail in a display of Tropical Fish Hobbyist publications. Having been a turtle lover since early childhood, Living Turtles of the World was a dream come true, written by some man named Peter Charles Howard Pritchard, from England no less. In addition to the worldwide coverage of turtles and the many pictures of species I never previously knew existed, I was drawn to the bibliography, which opened the door to a primary scientific literature that I also knew nothing about. That, of course, led me to Archie Carr's 1952 Handbook of Turtles, and since both men were at the University of Florida, my future was clear, and 3 years later I headed to UF for grad school.

But to my great disappointment, when I arrived, Peter had completed his doctorate and moved on to work with the Florida Audubon Society. And so I began my relationship with Peter the old-fashioned way—via snail mail. It was always a treat to get a letter from Peter, especially given the impressive and distinctive powder blue Florida Audubon stationery. Even his stationary was larger than life to me as a naive grad student, and I still have every letter Peter sent me.

As for so many others, Peter has been a lifelong inspiration to me. Among other things, we shared a deep appreciation of the value of the old scientific literature, something increasingly unappreciated today. His many monographs stand as tributes to the value of the works of those who have gone before. In 1159, John of Salisbury wrote, “Bernard of Chartres used to compare us to dwarfs perched on the shoulders of giants. He pointed out that we see more and farther than our predecessors, not because we have keener vision or greater height, but because we are lifted up and borne aloft on their gigantic stature.” Peter was truly one of those giants, both in stature and in intellect, and it was my privilege to call him a colleague and friend for 48 years.