Biology_of_Turtles

Figure 6.3  Mean kinematic profiles of femoral protraction and retraction (i.e., angle from the transverse plane) for (a) Trachemys scripta swimming, (b) Apalone spinifera swimming, (c) T. scripta walking, (d) A. spinifera walking. Each kinematic trial was normalized to the same duration, and values for kinematic angles were interpolated for 100 equally spaced increments through the locomotor cycle, allowing mean angles and standard errors to be calculated for each 1% increment through each cycle type in each species. Angle values

± 1 SE are plotted for every fourth increment (every 4% through the cycle); N = 10 trials for T. scripta swimming, 11 trials for T. scripta walking, and 13 trials for A. spinifera in both swimming and walking. Angles of 0° indicate that the femur is perpendicular to the anteroposterior midline of the turtle, with negative values indicating that the distal end of the femur is directed posteriorly and positive values indicating the distal end of the femur is directed anteriorly. The vertical line in each panel divides the cycle into recovery (R) and thrust

(T) phases for swimming, or swing (SW) and stance (ST) phases for walking.

In contrast to kinematic patterns for femoral motion at the hip, patterns of knee motion showed changes between water and land that were similar in both sliders and spiny softshells. In water, both species show a major phase of knee flexion during the recovery portion of the stroke, followed by a phase of knee extension that progresses into the thrust phase of the stroke (Figure 6.7A,B). Some brief periods of flexion may occur during the extension phase but these are minor excursions compared to the main phase of knee flexion at the beginning of the aquatic limb cycle. Terrestrial knee movements differ significantly from aquatic knee movements in both species. On land, both species show two cycles of knee flexion and extension: the first occurs during femoral protraction and the second occurs during femoral retraction (Figure 6.7C,D). Thus, during swing phase (femoral protraction) the knee initially flexes as the leg comes forward but then extends as the foot reaches out to contact the ground. After the foot reaches the ground, the knee flexes again at the beginning of limb retraction (as the limb begins to support the weight of the body) and then extends a second time as the limb pushes off the ground and propels the body forward. Both species show similar magnitudes of knee flexion during swing phase and stance phase. However, despite the similarities in adjustments of knee kinematics between water and land in these species, sliders and spiny softshells also showed differences in their knee kinematics. In particular, the overall range of angles

148

Table 6.1

Values (mean ± SD) of Kinematic Variables Measured from the Hindlimbs of Adult Slider (T. scripta) and Spiny Softshell (A. spinifera)

Turtles during Aquatic (Water) and Terrestrial (Land) Locomotion.

The P values reported are for Mann-Whitney U-tests of the following comparisons in each row: P1 = T. scripta, water vs. land; P2 = A. spinifera, water vs. land; P3 = water, T. scripta vs. A. spinifera; P4 = land, T. scripta vs. A. spinifera.

* Significant at P < 0.05.

Turtles of Biology

Figure 6.5  Mean kinematic profiles of femoral elevation and depression (i.e., angle from the horizontal plane) for (A) Trachemys scripta swimming, (B) Apalone spinifera swimming, (C) T. scripta walking, (D) A. spinifera walking. Figure format and method of profile calculation are the same as in Figure 6.3. Angles of 0° indicate a horizontal femur, with negative values indicating that the distal end of the femur is depressed below the proximal end, and positive values indicating the distal end of the femur is elevated above the proximal end.

C D

Figure 6.6  Still images from lateral view video footage of turtle hindlimbs during aquatic and terrestrial locomotion, comparing femoral depression between species. (A) Trachemys scripta, swimming, (B) Apalone spinifera swimming, (C) T. scripta walking, (D) A. spinifera walking. Scale = 1 cm. White lines highlight the position of the femur relative to the horizontal plane.

Figure 6.7  Mean kinematic profiles of knee flexion and extension for (A) Trachemys scripta swimming,

(B) Apalone spinifera swimming, (C) T. scripta walking, (D) A. spinifera walking. Figure format and method of profile calculation are the same as in Figure 6.3. Larger angles indicate greater knee extension; 180° would indicate a straight knee joint.

swimming and walking in T. scripta. PIFI activity extends through much of femoral protraction in both behaviors but its onset begins prior to the beginning of the recovery phase (Figure 6.9), suggesting its initial contraction might serve to brake limb retraction at the end of the thrust phase in anticipation of the limb protraction to occur in the following recovery phase. Flexor tibialis is active primarily during limb retraction, and during swimming in particular its activity does not substantially overlap with knee flexion, suggesting that, at least in swimming, its primary function may be limb retraction rather than knee flexion. However, unlike PIFI and flexor tibialis, femorotibialis shows a dramatic change in activation timing between swimming and walking with a second burst during terrestrial locomotion that corresponds to the second phase of knee extension exhibited by this species while walking on land (Figure 6.9). Although only femorotibialis showed dramatic differences in burst timing among the muscles examined in T. scripta, all three muscles examined show significant differences in burst intensity between habitats (Mann-Whitney U-tests, P < 0.05), with PIFI showing higher amplitude bursts during walking, and femorotibialis and flexor tibialis showing higher amplitude bursts during swimming (Figure 6.9).

EMG patterns for A. spinifera show several similarities to those for T. scripta. As in T. scripta, in A. spinifera activity of PIFI begins before the start of limb protraction and continues through much of the duration of limb protraction with little change in burst timing between swimming and walking (Figure 6.9). Also similar to T. scripta, the single flexor tibialis burst in A. spinifera occurs later in the locomotor cycle during swimming than during walking, coinciding with limb retraction, rather than knee flexion, during aquatic locomotion. Both PIFI and flexor tibialis show minor differences in burst timing between the species (e.g., PIFI bursts onset earlier before the start of limb protraction in T. scripta than in A. spinifera); however, these small differences do not suggest substantial changes in the functional roles of these muscles between these species.

Although T. scripta and A. spinifera show many similarities in their hindlimb motor patterns, significant differences also emerge between these species. As in T. scripta, in A. spinifera the

Figure 6.8  Mean kinematic profiles of paddle (i.e., foot) orientation during swimming for (A) Trachemys scripta hindfoot, (B) Apalone spinifera hindfoot, (C) T. scripta forefoot, (D) A. spinifera forefoot. Figure format and method of profile calculation are the same as in Figure 6.3. The angle plotted is the angle between a vector pointing forward along the anteroposterior midline of the turtle and a vector emerging from the plantar surface of a plane defined by the ankle (hindfoot) or wrist (forefoot) and the tips of digits 1 and 5, transformed by subtracting 90° from each value. Thus, a low-drag orientation of a foot (ideal feathering) is indicated by an angle of 0°, whereas a high-drag orientation (with increased thrust-generating potential during limb retraction) is indicated by an angle of 90°. Data for forefeet were replotted for comparison from Pace et al. (2001).

femorotibialis exhibits a single burst of activity during swimming but two bursts of activity during walking, each of which correspond to one of the two phases of knee extension during terrestrial locomotion (Figure 6.9). A. spinifera femorotibialis EMG bursts also show greater mean amplitudes during swimming than during walking, as in T. scripta (Mann-Whitney U-test, P < 0.05). However, during swimming in T. scripta the femorotibialis burst occurs during the recovery phase of the stroke, whereas in swimming A. spinifera this burst occurs during the thrust phase of the stroke (Figure 6.9). A further distinction in motor patterns between the species is that, unlike T. scripta, in A. spinifera mean EMG burst amplitudes do not differ between water and land for either flexor tibialis or PIFI (Figure 6.9). In addition, differences in terrestrial knee kinematics between sliders and spiny softshells suggest the potential for a further contrast in hindlimb muscle function between these species. Because spiny softshells walk with the knee more extended than it is in sliders, the femorotibialis muscle of A. spinifera could be contracting when it is already highly shortened. As a result, in aquatic specialist softshell turtles, the femorotibialis might be active at a length that is suboptimal for force generation (Schmidt-Nielsen, 1990).

6.2.3Conclusions from Comparisons of Hindlimb Function in Adult Slider

and Softshell Turtles: Functional Consequences of Habitat Specialization

Sliders and spiny softshells show largely similar ways of adjusting limb function between aquatic and terrestrial habitats. These adjustments can be widely distributed in the locomotor system (Gillis & Blob, 2001), though they are not necessarily present among all muscles (e.g., A. spinifera PIFI).

Figure 6.9  Representative mean activity patterns for three focal hindlimb muscles during swimming (shaded bars) and walking (unshaded bars) for Trachemys scripta (left column) and Apalone spinifera (right column), synchronized with mean kinematic profiles of knee kinematics for both behaviors in each species (replotted from Figure 6.7). Within a muscle, greater bar thickness in swimming or walking indicates significantly greater mean EMG amplitude for that behavior (Mann–Whitney U-test, P < 0.05). Vertical lines demarcate cycles into recovery (R) and thrust (T) phases for swimming and swing (SW) and stance (ST) phases for walking; these demarcations are both indicated with appropriate labels on muscle activity graphs. All error bars indicate 1 SE.

However, even with largely similar modulation patterns between T. scripta and A. spinifera, some prominent contrasts in hindlimb function can be identified between these species. For example, sliders depress the femur twice as much as softshells during both swimming and walking. Sliders also flex their knee significantly more than softshells during walking (i.e., softshells walk with straighter hindlegs). In addition, although the femorotibialis is active during knee extension in both species, during swimming this activity occurs during the recovery phase of the stroke in sliders but during the thrust phase of the stroke in spiny softshells. A. spinifera also do not show all of the changes in burst amplitude between swimming and walking that are exhibited by T. scripta.

The changes in motor function between swimming and walking exhibited by sliders and softshell turtles provide some broader insights into how muscular function can be modulated to perform a variety of tasks. First, both activation timing and intensity (e.g., EMG burst amplitude) can be modulated independently between behaviors (Gillis & Blob, 2001), and these patterns may differ for a muscle even between closely related species. For example, in femorotibialis both the timing

locomotion soon after hatching (Cagle, 1950; Miller et al., 1987; Zani & Claussen, 1994). Even for young showing coordinated locomotion, patterns of hindlimb function might be similar across species of juvenile turtles for two primary reasons. First, hatchling turtles are typically less than 25% adult carapace length. Because small animals tend to use high frequency limb movements (Heglund & Taylor, 1988), there may be a limited number of ways for juvenile turtles to move their limbs very quickly (Carrier, 1996), causing all juvenile turtles to use similar motor control patterns. Second, for T. scripta and A. spinifera in particular, juveniles are more similar in habits than adults. Juveniles of both sliders and spiny softshells are generally predators of evasive aquatic prey (typically small invertebrates); though A. spinifera maintain predatory habits into adulthood, T. scripta tend to shift to a diet dominated by aquatic plants as they grow larger in size (Cagle, 1950; Dalrymple, 1977; Hart, 1983). Thus, differences in locomotion among adults that could be associated with the requirements for acquiring different types of food (evasive versus non-evasive) might not be expected among juveniles that face similar locomotory demands.

6.3.1Experimental Methods

To test if contrasts in limb function between adults of different turtle species can be traced to their juveniles, we collected high-speed video and hindlimb EMGs during swimming and walking from juvenile T. scripta and A. spinifera (8 to 15 weeks posthatching, 18 to 25% adult length, N = 2 individuals per species) purchased from a commercial supplier (Turtletown, Waukegan, Illinois, U.S.A.). Patterns observed in juveniles were then compared to those exhibited by adults. Methods of video and EMG data collection were the same as those described previously for adults of these species and targeted the same three muscles (femorotibialis, flexor tibialis complex, and PIFI), with the exception that filming was frequently conducted at 125 Hz rather than 60 Hz. Kinematic analyses (digitizing of limb movements, smoothing of kinematic variables) and EMG analyses (calculation of burst variables) also proceeded as described previously for adults, standardizing procedures to facilitate comparisons between the groups.

6.3.2Results: Ontogeny of Hindlimb Function in Slider and Softshell Turtles

Limb cycle frequencies in juvenile sliders and softshells for both swimming (T. scripta: 3.81 ± 0.13 Hz; A. spinifera: 3.00 ± 0.12 Hz; mean ± SE) and walking (T. scripta: 2.47 ± 0.22 Hz; A. spinifera:

Figure 6.10  Comparison of locomotor cycle frequencies for juvenile versus adult turtles during swimming (shaded bars) and walking (unshaded bars) for (a) Trachemys scripta and (b) Apalone spinifera. Data for adults replotted from Figure 6.2; N = 10 cycles for T. scripta juveniles in both swimming and walking, 7 cycles for A. spinifera juveniles in swimming, and 5 cycles for A. spinifera juveniles in walking. Error bars indicate 1 SE. *All comparisons of juvenile to adult of the same species for locomotion within a specific environment are significantly different (Mann–Whitney U-tests, P < 0.01).

Percentage of Locomotor Cycle

Figure 6.11  Comparison of mean kinematic profiles for femoral depression during swimming in juvenile

(A) Trachemys scripta and (B) Apalone spinifera. Lateral view still images from high-speed video recordings accompany each plot; arrows indicate location of the femur. Format for kinematic plots follows that in Figure 6.3. Error bars indicate 1 SE. Scale = 1 cm.

1.85 ± 0.09 Hz; mean ± SE) are significantly higher than those for adults of both of these species (Mann–Whitney U-tests, P < 0.01; Figure 6.10). This result supports the possibility that juvenile turtles of different species might show similarities of hindlimb function related to their shared small size.

Kinematic comparisons of juvenile sliders and spiny softshells indicate that some of the main contrasts in hindlimb function found between adults of these species are not evident early in ontogeny. For example, in contrast to patterns observed in adults, juvenile A. spinifera actually depress the femur more (−26.4° ± 6.6°) than juvenile T. scripta (−17.2° ± 1.3°) during swimming (MannWhitney U-test, P < 0.01; Figure 6.11; Table 6.2), potentially increasing the upward component of thrust in A. spinifera. Similarly, in contrast to patterns observed among adults, juvenile A. spinifera flex the knee more (74.1° ± 10.9°) than juvenile T. scripta (93.4° ± 14.2°) during terrestrial walking (Mann-Whitney U-test, P < 0.01; Figure 6.12; Table 6.2).

Comparisons of EMG burst patterns between juvenile T. scripta and A. spinifera show that a major contrast in muscle activation between adults of these species is present in juveniles: as in adults, the femorotibialis is active during the recovery phase of swimming in juvenile sliders but during the thrust phase of swimming in juvenile spiny softshells (Figure 6.13). However, juveniles of these species also show a number of contrasts in motor patterns beyond those seen in adults. For example, in both species femorotibialis shows one EMG burst during swimming cycles in adults but shows two EMG bursts during swimming cycles in juveniles (Figure 6.13). In addition, despite the similar EMG burst patterns shown by PIFI both between species and between locomotor environments in adults, juvenile T. scripta and A. spinifera show very different PIFI burst patterns both between species and between water and land (Figure 6.13).