Table 1.
Mean fractional anisotropy (FA), whole brain volume (WBV), percentage of whole brain volume contained in inferior colliculi (IC/ WBV) and percentage of whole brain volume contained in cerebella (Cerebella/ WBV) for each specimen.
Fig 1.
a. Overview of ROIs and waypoint/exclusion criteria for each trace.
In all traces, the inferior colliculi (IC) were the primary seed/ROI through which pathways had to pass. Ascending auditory traces simply show the results of a plain IC-seed protocol, and are typically shown/discussed for their qualitative information they communicate about the brain regions that connect to IC. IC-ipsilateral cortex is the corresponding trace used for quantification of how many IC pathways project to the ipsilateral cortex on each side; any pathways that traverse cerebellum or cross the midline were excluded. IC-contralateral cortex traces included only pathways that crossed the seeded IC and the midline but did not traverse the cerebellum, and are presented in both quantitative and qualitative results. IC-contralateral cerebellum traces included pathways that traversed IC and the cerebellum contralateral to the seeded IC, and are also presented in both qualitative and quantitative results. Green signifies that a region was used as a seed or waypoint, i.e., pathways had to contact them in order to be included, while red signifies that a region was excluded, i.e., pathways that contacted them were excluded. Only left-side traces are shown in this figure, but each of these protocols were performed for both right and left sides. b. Tract strength and laterality index. Panel (a) contains bar graph representations of the corrected tract strength of each trace in each specimen. See Table 2 for exact values. Panel (b) contains graphical representations of the laterality index (LI) of each trace in each specimen (LI formula and LI graph design adapted from Wright et al. [42]). LI values range from -1, maximally right-lateralized, to 1, maximally left-lateralized, and are given to the right of each graph. Refer to Fig 1a for details on the ROIs and criteria for each trace. IC = inferior colliculi.
Table 2.
Tract strength and laterality factor. Tract strength (i.e., waytotal divided by whole brain volume) and lateralization factor (i.e., how many times more tracts were observed in one hemisphere versus the other) for each trace in each specimen. Ipsi = ipsilateral, Contra = contralateral, IC = inferior colliculi, and LF = lateralization factor. Right versus left designations denote which cortical or cerebellar hemisphere was targeted more strongly by a respective inferior colliculus, rather than which the side of the IC was seeded. Refer to Fig 1a for details on the ROIs and criteria for each trace.
Fig 2.
Cortical and cerebellar targets of inferior colliculi.
Panels (a)-(d) contain 3D representations of ascending auditory tracts for each specimen. Tracts seeded in the right inferior colliculi (IC) are colored red, while those seeded in the left are colored blue. Adjacent to the 3D view, bright green boxes display coronal cross-section views of these traces in each specimen, highlighting their path through the ventral thalamus (putative medial geniculate nucleus; for a more detailed view see Fig 3.) Panels (e)-(l) contain orthographic views of some primary cerebellar targets of traces between the IC and cerebella in each specimen. Tracts targeting the right cerebellum are colored orange, while those targeting the left are colored turquoise. For each specimen, two orthographic-view images are given: one highlighting a prominent projection site in the left cerebellum, the other highlighting a prominent projection site in the right. IC = inferior colliculi, Cb = cerebellum. For more details on cerebellar projection sites, see Fig 6, S9 Table, S1 Text, S6 Figures or S7 Figures in S1 File. In all panels, the correspondence between the number of predicted tracts and the colors displayed is given in the gradient color bar on the right edge, with brighter colors indicating a higher number of tracts predicted in a given voxel and vice versa.
Fig 3.
Ascending auditory pathways transit putative MGN.
Coronal and axial cross-sections depicting the ventral thalamic transit of ascending auditory pathway traces in B. borealis (a, b) and L. acutus (c, d). Cross section images without (a, c) and with (b, d) tracts are shown; for B. borealis (a, b), a high-resolution T1-weighted image is shown, while a low-resolution b0 image is shown for L. acutus (c, d). Tracts are thresholded with minimum and maximum values of 0.1% and 5% of the trace waytotal, respectively, in B. borealis (b) and with minimum and maximum values of 1% and 30% of the trace waytotal, respectively, in L. acutus (d). The correspondence between predicted number of tracts in a given voxel and displayed colors of tracts is given in the gradient color bar on the right edge of panels b and d. Bright green boxes mark putative medial geniculate nucleus in each panel. MGN = medial geniculate nucleus and IC = inferior colliculi.
Fig 4.
Cortical targets of ascending auditory traces in odontocetes.
Orthographic cross-sections displaying key cortical targets of the inferior colliculi in the odontocete specimens. Each trace is displayed with minimum and maximum values set to 0.1% and 5% of its waytotal, respectively. In all panels, traces from right IC are colored red and those from left IC are colored blue. Gradient color bars specify the correspondence between predicted number of tracts in a given voxel and displayed colors of tracts. Panel (a) displays coronal and axial views of projections to lateral extremes of the left and right cortex in D. delphis and L. acutus, respectively. Panel (b) displays axial and coronal views of tracts transiting the basal ganglia, potentially caudate nucleus, in all odontocetes. Panel (c) displays axial and coronal views of continuous projections between both colliculi and a dorsal-rostral region of left cortex in D. delphis. Finally, panel (d) displays sagittal, coronal, and axial views of projections to a caudal, ventral, and medial cortical area in all odontocetes. Bright green boxes mark the notable cortical projection sites mentioned in-text.
Fig 5.
Rotating 3-dimensional tractograms of ascending auditory pathways in all specimens. a. D. delphis. b. S. attenuata. c. L. acutus. d. B. borealis. a-d. All displayed tractograms were seeded in the IC (refer to Fig 9 or S3 Figures in S1 File to view the masks of IC that served as seeds). These traces are displayed with minimum and maximum values set to 0.1% and 5% of their waytotals, respectively. In all panels, traces from right IC are colored red and those from left IC are colored blue. Gradient color bars specify the correspondence between predicted number of tracts in a given voxel and displayed colors of tracts.
Fig 6.
Color-coded diagram displaying cerebellar lobules targeted in the left, right, or bilateral cerebella in odontocete and mysticete specimens.
Within the central venn diagram, superscript “M” signifies that a subregion was targeted in the mysticete, while “O” signifies a subregion’s targeting by one or more odontocetes. Subregions that were targeted in the left cerebellar hemisphere are placed in the blue circle on the left, while right-hemisphere subregions are placed in the red circle on the right, and bilaterally-targeted subregions are placed in the overlap region colored purple. The outer ring of the central venn diagram specifies the functions associated with each lobule in other mammals [35,83–95]. Each function is placed in proximity with the lobule to which it corresponds in the inner venn diagram, and is color-coded in accordance with the subregion to which it corresponds as well. See the discussion subsection Specific cerebellar targets for more details on correspondences between subregions and functions. In the bottom corner, a smaller venn diagram displays the subregions only targeted in odontocetes, and uses the same color-coding scheme to denote whether these regions were targeted on the left side, right side, or bilaterally.
Fig 7.
Subcortical features of IC-cerebellar traces in odontocetes.
Orthographic cross-sections displaying key subcortical features of the traces between IC and cerebellum in the odontocete specimens (a) D. delphis, (b) S. attenuata, and (c) L. acutus. Each trace is displayed with minimum and maximum values set to 0.1% and 5% of its waytotal, respectively. Traces between right IC and left cerebellum are colored blue, while those between left IC and right cerebellum are orange. The correspondence between predicted number of tracts in a given voxel and displayed colors of tracts is given in the gradient color bars on the right edges of the panels. IC = inferior colliculi and Cb = cerebellum.
Fig 8.
Putative model of how differential auditory cortical-cerebellar connectivity may relate to differential mechanisms and functions of hearing and sound production in odontocetes versus mysticetes. Red arrows color-code for right-cerebellar lateralization of IC-cerebellum tracts in odontocetes, while blue arrows color-code for left-cerebellar lateralization of IC-cerebellum tracts in mysticetes. The solid red arrow is thicker than the blue one to represent the higher tract strength of IC-cerebellar pathways in odontocetes. Solid red and blue arrows reflect pathways that were robustly demonstrated (i.e., withstood rigorous thresholding) in this study, while dashed red and blue arrows represent tracts that were observed more weakly (i.e., did not withstand rigorous thresholding), but may be directly or indirectly relevant to sound production pathways, and thus merit further investigation. Grey dashed arrows represent pathways that were not directly measured or observed, or could not be measured in this study (i.e., peripheral connections), but are widely conserved and/or canonical circuits between the ROIs that may be relevant to sound production. The grey dashed arrow that connects M1 and cerebellum but does not transit IC in the mysticete is unique in that it represents a potential takeaway from our negative finding here: since very few projections from IC-cerebellum reach putative M1 in the mysticete, it is likely that more general, widely conserved, and canonical M1-cerebellum tracts are present in these brains and involved in complex sound production [98]- though they may not be as densely connected to subcortical auditory centers as they are in echolocating odontocetes. Because our DTI tractographical method does not provide an index of directionality (i.e., efferent versus afferent, input versus output), and because our observed pathways likely involve cortico-cerebellar and cortico-cortical feedback loops, all arrows representing observed streamlines are bidirectional. Brainstem cross-sections display the hypothesized brainstem nuclei relevant for sound production in each suborder, and the hypothesized peripheral phonatory organs that would in turn receive motor signals from these nuclei are placed underneath. Coronal cortical cross-section adapted from Brain Catalogue [99], and brainstem cross-section adapted from Neurosurgical Atlas [100].
Fig 9.
Inferior Colliculi and Cerebellum ROIs.
Sagittal, coronal, and axial cross-sections showing the masks made for the cerebella (turquoise and yellow for left and right, respectively) and inferior colliculi (blue and red for left and right, respectively) in B. borealis. Refer to S3 Figures in S1 File to view these masked ROIs in the other specimens.