Evolutionary novelty versus exaptation: oral kinematics in feeding versus climbing in the waterfall-climbing Hawaiian Goby Sicyopterus stimpsoni

Abstract

Species exposed to extreme environments often exhibit distinctive traits that help meet the demands of such habitats. Such traits could evolve independently, but under intense selective pressures of extreme environments some existing structures or behaviors might be coopted to meet specialized demands, evolving via the process of exaptation. We evaluated the potential for exaptation to have operated in the evolution of novel behaviors of the waterfall-climbing gobiid fish genus Sicyopterus. These fish use an "inching" behavior to climb waterfalls, in which an oral sucker is cyclically protruded and attached to the climbing surface. They also exhibit a distinctive feeding behavior, in which the premaxilla is cyclically protruded to scrape diatoms from the substrate. Given the similarity of these patterns, we hypothesized that one might have been coopted from the other. To evaluate this, we filmed climbing and feeding in Sicyopterus stimpsoni from Hawai'i, and measured oral kinematics for two comparisons. First, we compared feeding kinematics of S. stimpsoni with those for two suction feeding gobiids (Awaous guamensis and Lentipes concolor), assessing what novel jaw movements were required for algal grazing. Second, we quantified the similarity of oral kinematics between feeding and climbing in S. stimpsoni, evaluating the potential for either to represent an exaptation from the other. Premaxillary movements showed the greatest differences between scraping and suction feeding taxa. Between feeding and climbing, overall profiles of oral kinematics matched closely for most variables in S. stimpsoni, with only a few showing significant differences in maximum values. Although current data cannot resolve whether oral movements for climbing were coopted from feeding, or feeding movements coopted from climbing, similarities between feeding and climbing kinematics in S. stimpsoni are consistent with evidence of exaptation, with modifications, between these behaviors. Such comparisons can provide insight into the evolutionary mechanisms facilitating exploitation of extreme habitats.

Figure 1. Still images of S. stimpsoni in (a) ventral and (b) lateral views, illustrating anatomical landmarks that were digitized to generate kinematic data.

For ventral view (a), labeled points are as follows: (1) anterior edge of upper lip, (2) anterior tip of inner edge of upper lip, (3) anterior tip of mandibular symphysis, (4) right caudo-lateral tip of mouth, (5) left caudo-lateral tip of mouth, (6) midpoint on right side of mandible between mandibular symphysis and right caudo-lateral tip, (7) midpoint on left side of mandible between mandibular symphysis and left caudo-lateral tip, (8) hyoid arch, (9) midline joint between left and right branchiostegal rays, (10) caudolateral margin of right operculum, (11) caudolateral margin of left operculum, (12) right pectoral fin base, (13) left pectoral fin base, and (14) anterior tip of pelvic sucker. For lateral view (b), labeled points are as follows: (15) anterior tip of upper lip, (16) anterior edge of upper lip base, (17) caudal tip of junction between maxilla and dentary, (18) anterior edge of neurocranium, (19) center of eye, (20) junction between neurocranium and epaxial muscle insertion, (21) caudal edge of operculum, (22) dorsal edge of pectoral fin base, and (23) ventral edge of pectoral fin base spine.

Figure 2. Geometric model for the calculation of oral sucker area from digitized landmarks in ventral view footage of feeding and climbing by Sicyopterus stimpsoni.

(a) Outline sketch of the mouth of S. stimpsoni in ventral view, with superimposed greyscale shaded geometric shapes defined by labeled digitized landmarks. A, digitized Point 1 (anterior edge of upper lip); B, digitized Point 3 (anterior tip of mandibular symphysis); C, digitized Point 4 (right caudolateral tip of mouth); D, digitized Point 6 (midpoint of mandible on the right side); E, calculated midpoint between points B and C; θ, angle between vectors AB and AC. Note that digitized points are shown as open circles, and calculated points are shown as solid circles. (b) Geometric shapes from (a) separated into three groups (designated X, Y, and Z), with formulae for calculation of their areas. Sucker area was modeled as the sum of these three geometric areas, which assume symmetry between left and right sides.

Figure 3. Representative lateral and ventral view still frames from high-speed video of (a) feeding and (b) climbing cycles of Sicyopterus stimpsoni.

Panels are sequential from top to bottom for each behavior, with elapsed time through the cycle reported in lateral frames. Note in (b) that the fish climbs upwards (toward the top of each frame) as frames are viewed in order from top to bottom. Because climbing cycles are longer in duration than feeding cycles, the five time points illustrated for each behavior represent equivalent fractions of time through the kinematic cycle, at 0%, 25%, 50%, 70%, and 100%. All scale bars equal 1 cm. Note that lateral and ventral views for each behavior are filmed at different magnifications.

Figure 4. Comparative profiles of cranial kinematics for Sicyopterus stimpsoni during feeding (solid triangles) and climbing (open circles) behaviors.

Descriptions of the calculation of each variable are provided in the text. Plots show mean (± s.e.m. values of each variable, averaged across all cycles for each behavior for every 5% increment of cycle duration. (a) Cranial elevation angle, (b) premaxillary protrusion angle, (c) premaxillary protrusion length, (d) hyoid retraction angle, (e) hyoid retraction length, (f) mandibular retraction length, (g) opercular expansion length, and (h) oral sucker area. All linear measurements are normalized by body length (BL), or BL² for oral sucker area.