Selective binocular vision loss in two subterranean caviomorph rodents: Spalacopus cyanus and Ctenomys talarum

Abstract

To what extent can the mammalian visual system be shaped by visual behavior? Here we analyze the shape of the visual fields, the densities and distribution of cells in the retinal ganglion-cell layer and the organization of the visual projections in two species of facultative non-strictly subterranean rodents, Spalacopus cyanus and Ctenomys talarum, aiming to compare these traits with those of phylogenetically closely related species possessing contrasting diurnal/nocturnal visual habits. S. cyanus shows a definite zone of frontal binocular overlap and a corresponding area centralis, but a highly reduced amount of ipsilateral retinal projections. The situation in C. talarum is more extreme as it lacks of a fronto-ventral area of binocular superposition, has no recognizable area centralis and shows no ipsilateral retinal projections except to the suprachiasmatic nucleus. In both species, the extension of the monocular visual field and of the dorsal region of binocular overlap as well as the whole set of contralateral visual projections, appear well-developed. We conclude that these subterranean rodents exhibit, paradoxically, diurnal instead of nocturnal visual specializations, but at the same time suffer a specific regression of the anatomical substrate for stereopsis. We discuss these findings in light of the visual ecology of subterranean lifestyles.

Figure 1

(a–c) Perspective views of orthographic projections of the frontal binocular (blue, a and b) and caudal monocular (brown, c and d) visual fields in the subterranean S. cyanus and C. talarum. In both rodents the maximum binocular overlap is located between 25–50° in elevation (a,b). Note the bottleneck between −10° and −20° of the binocular overlap in S. cyanus resulting from the nose protruding into the binocular field (a). Interestingly, this bottleneck shape of the binocular field is absent in C. talarum (b). The diagrams use the conventional latitude and longitude system. The drawings of a representative subterranean rodent inside the orthographic projections, courtesy of Ricardo Di Parodi.

Figure 2. Retinal whole mounts.

Topographic isodensity map reconstructions of cells in the retinal ganglion cell layer (GCL) of S. cyanus (a) and C. talarum (b). The black area in the central retina indicates the optic nerve head. Note that the decline in cell density is more pronounced towards the dorsal retina. (c–f) Insets with red and blue contours showing photomicrographs of Nissl-stained regions of the dorsal and temporal retina, respectively, as indicated in the wholemounts. Arrowheads indicate neurons in the ganglion cell layer. D = dorsal; N = nasal; T = temporal; V = ventral. Scale bar insets = 20 μm.

Figure 3

Contralateral sagittal sections showing CTB-labeled retinal fibers counterstained with Giemsa in S. cyanus and C. talarum (a,b). N. geniculatus lateralis pars ventralis (GLv), n. geniculatus lateralis pars dorsalis (GLd), superior colliculus (SC), pretectum (PRT), medial terminal nucleus (MTN), suprachiasmatic nucleus (SCN), optic tract (Ot), optic chiasma (CO). Rostral is to the left. Scale bar = 0.8 cm.

Figure 4

Sagittal sections showing CTB-labeled retinal terminals in the contra- and ipsilateral GLd of S. cyanus (a,c) and C. talarum (b,d). Arrowheads in (c) indicate ipsilateral terminals in the GLd of S. cyanus. Note that in (d) there are no retinal fibers in the ipsilateral GLd of C. talarum. Insets in (a,b) represent low magnification photomicrographs of the respective sagittal GLd. Counterstaining with Giemsa. R = rostral, C = caudal, V = ventral. Scale bar = 1 mm.

Figure 5

Transverse sections showing CTB-Alexa (488 and 546) labeled retinal fibers in the optic chiasma (CO) and optic tract (Ot) (a,b). Left (CTb-Alexa 488) and right (CTb-Alexa 546) labeled retinal terminals in the n. geniculatus lateralis pars dorsalis (GLd) of C. talarum (c,d). Note that there is no labeling of ipsilateral fibers in the GLd. Orientation in (c) applies to all sections. D = dorsal; V = ventral.

Figure 6

Sagittal and coronal sections showing CTB-labeled retinal fibers in the contra- and ipsilateral SC of S. cyanus (a,c,e) and C. talarum (b,d,f). Note that there are no labeled retinal fibers in the ipsilateral SC of C. talarum. Insets in (a,b) represent low magnification photomicrographs of the respective sagittal and coronal SC. Arrowheads in (c,e) indicate retinal terminals in the SO of S. cyanus. Counterstaining with Giemsa. R = rostral, C = caudal, V = ventral, D = dorsal.

Figure 7

Transverse section showing CTB-Alexa labeled retinal fibers in the left (Alexa-488) and right (Alexa 546) superior colliculus of C. talarum (a). Note that there is no labeling of ipsilateral fibers in the SC, confirming the complete loss of this projection in the facultative non-strictly subterranean C.talarum. D = dorsal; V = ventral.

Figure 8

(a) Plot of the contralateral GLd volume relative to the brain volume in C. talarum and S. cyanus. No significant differences were observed between species (Kruskal-Wallis and Mann Whitney U-tests; α = 0.05; P > 0.05). Relative volume values are in multiples of 10⁻⁷. (b) Plot of the contralateral SC volume relative to the brain volume (Y axis). Asterisk indicates a significant difference between groups (Kruskal-Wallis and Mann Whitney U-tests; α = 0.05; P = 0.05). Relative volume values are in multiples of 10⁻⁶. The solid lines contained within the yellow boxes in (a,b) represents the mean.

Figure 9. Plot of the ipsilateral GLd volume relative to the brain volume in O. lunatus (nocturnal, surface-dwelling), O. degus (diurnal, surface-dwelling), S. cyanus (diurnal, subterranean) and C. talarum (diurnal, subterranean).

Differences between groups are significant (Kruskal-Wallis and Mann Whitney U-tests; α = 0.05; P = 0.05). Note that the value for C. talarum is 0. Relative volume values are in multiples of 10⁻⁸. The solid lines contained within the yellow boxes represent the mean.