Since the 1960s, the California sea lion (Zalophus californianus) has been a model organism for sensory, communication, and cognition studies in marine mammals. Behavioral research on this animal has focused on visual acuity and discrimination between objects (Kastak and Schusterman, 2002a; Schusterman et al., 1965; Schusterman and Balliet, 1970), hearing sensitivity (Schusterman, 1974; Reichmuth et al., 2007), vocalizations and communication (Schusterman et al., 1966; Schusterman and Balliet, 1969; Hanggi and Schusterman, 1990; Gisiner and Schusterman, 1991), and memory (Schusterman et al., 1993; Kastak and Schusterman, 2002a). However, to the best of our knowledge, no formal studies have focused on the anatomy and size of brain structures in this species.
In the wild, California sea lions can be exposed to domoic acid (DA), a natural marine neurotoxin produced by diatoms belonging to the Pseudo-nitzschia genus. DA acts as an excitotoxin by binding to the α-amino-5-hydroxy-3-methyl-4-isoxazole propionic acid and kainate subclasses of neuronal glutamate ionotropic receptors (Jeffrey et al., 2004). This binding causes massive cell depolarization, resulting in dysfunction and death of cells expressing these receptors (Olney et al., 1979). In 1998, more than 400 sea lions were exposed to DA through contaminated prey (Scholin et al., 2000). Sea lions that died acutely and contained detectable levels of DA in blood and urine exhibited lesions in the limbic system (Scholin et al., 2000). These lesions were characterized by neuronal necrosis in the hippocampal formation, specifically granule cells in the dentate gyrus and pyramidal cells in sectors CA4, CA3, and CA1 of the cornu ammonis (Silvagni et al., 2005).
A second novel neurological syndrome has been identified in California sea lions associated with chronic exposure to sublethal levels of DA (Goldstein et al., 2008). Sea lions examined alive by magnetic resonance imaging (MRI) exhibited varying degrees of hippocampal atrophy, thinning of the parahippocampal gyrus, and increases in the size of the temporal horn of the lateral ventricle (Goldstein et al., 2008). This examination was based on subjective analysis and not on volumetric measurements. Now, there is concern that low levels of DA exposure in the developing fetus and neonate, via concentration in the amniotic fluid and mammary glands, respectively, may cause subtle changes in the brain that may result in long-term cognitive impairment (as reviewed by Ramsdell and Zabka, 2008). As humans can also be exposed to DA in seafood, there is a need to understand the range of effects that DA may have on naturally exposed mammals, so effects on humans can be predicted and prevented.
With the widespread exposure of wild sea lions to DA, there is a need to learn more about the normal brain of the California sea lion to accurately identify changes in the brain due to DA toxicity. MRI has been used recently to study the anatomy of cetacean brains that were removed from the skull and formalin fixed (Marino et al., 2001a, b, c, 2003a, b, 2004a, b) and that were freshly intact within the skull and attached to the body (Montie et al., 2007, 2008). For the Atlantic white-sided dolphin (Lagenorhynchus acutus), MRI has been used to acquire images of the brain to calculate the volumes of white matter (WM), gray matter (GM), cerebellum WM and GM, and the hippocampus at different developmental stages (Montie et al., 2008). Using these methods, MRI provides a means to evaluate normal brain structure and determine the size of brain regions in live California sea lions. This approach would be a valuable tool in assessing the effects of DA neurotoxicity in exposed animals.
Our goal in this study was to create a neuroanatomical atlas and establish a quantitative approach to determine the size of brain structures from MR images of a live California sea lion. Specifically, the objectives were to (a) present an anatomically labeled MRI-based atlas of a neurologically normal brain; (b) provide detailed labeling of hippocampal structures; (c) determine the WM and GM volumes of the total brain, cerebellum and brainstem combined, and cerebral hemispheres; and (d) determine the volumes of the left and right hippocampus and associated structures.
MATERIALS AND METHODS
The female California sea lion used in this study (Accession number = CSL7775; Name = “Kirina”) was rescued on 25 July, 2008 by the Marine Mammal Center, Sausalito, CA from Pico Point, San Simeon, CA (35°36′46.7994″N, −121°9′7.2″W) due to a fractured right hind flipper. Hematological and serum biochemical parameters were within the normal range for this species, other than an elevated white blood cell count. A neurological assessment (posture, mentation, body movements other than use of the right hind flipper, cranial nerve reflexes, and responsiveness to audible stimuli and visual approach) appeared normal. At the time of MRI, the body length was 117 cm, and the weight was 34.5 kg. These measurements along with the stage of tooth development indicated that the animal was approximately 1 year old (Greig et al., 2005). Radiography and MRI were used to evaluate the flipper and brain of the sea lion to determine its prognosis. MR examination did not reveal any brain pathologies; the brain appeared grossly normal.
Magnetic Resonance Data Acquisition
Radiographs of the whole body nullified any suspicion of metallic foreign bodies present in the animal. The animal was imaged at IAMs Pet Imaging Center, San Francisco, CA, following anesthesia with isoflurane. MRI scanning of the brain was completed with a 1.5-T Siemens Magnetom Symphony scanner (Siemens, Munich, Germany) equipped with a CP Extremity Coil. Following the localizer scan, T1-weighted images in the sagittal plane were acquired using a spoiled gradient echo (FLASH) sequence with the following parameters: TR = 22 ms; TE = 10 ms; FOV = 200 × 200 mm; slice thickness = 1 mm; and voxel size = 0.3 × 0.3 × 1 mm. Two-dimensional proton density (PD) and T2-weighted images in the transverse plane were acquired using a turbo spin-echo (TSE) sequence with the following parameters: TR = 3650 ms and TE = 14/98 ms for PD and T2, respectively; slice thickness = 2.5 mm; FOV = 150 × 150 mm; and voxel size = 0.3 × 0.3 × 2.5 mm. Additionally, two-dimensional PD- and T2-weighted images in the oblique plane (i.e., perpendicular to the long axis of the sylvian fissure and temporal lobe) were acquired using a TSE sequence with the following parameters: TR = 5470 ms and TE = 14/98 ms for PD and T2, respectively; slice thickness = 2.5 mm; FOV = 160 × 160 mm; and voxel size = 0.3 × 0.3 × 2.5 mm. The oblique orientation was selected to optimize viewing of the hippocampus, as previously described (Goldstein et al., 2008).
After radiography and MR examination, the animal was euthanized because of poor prognosis for rehabilitation and release due to the severity of osteomyelitis in the hind flipper. A necropsy was completed. The brain was removed, weighed, and archived whole in 10% neutral buffered formalin.
Anatomic Labeling and Nomenclature
Anatomical structures were identified using the brain atlas of the domestic dog (Beagle, Canis familiaris) (Dua-Sharma et al., 1970) and human (Nolte and Angevine, 2000) and labeled using nomenclature adopted from the English translation of Nomina Anatomica Veterinaria (ICVGAN, 2005) (Table 1). Comparisons of the California sea lion brain to other species in the Order Carnivora were made using the Comparative Mammalian Brain Collection website (http://www.brainmuseum.org/index.html) prepared by the University of Wisconsin, Michigan State University and the National Museum of Health and Medicine (Welker et al., 2009). Segmentation (i.e., assigning pixels to particular structures) and three-dimensional (3D) reconstructions of the brain surface were performed using the software program AMIRA 4.1.1 (Mercury Computer Systems, San Diego, CA). To create a 3D reconstruction of the brain surface, the pixels in the native (i.e., no image processing) T1-weighted images from the FLASH sequence were selected manually and defined as cerebrum, cerebellum, spinal cord, olfactory bulb and tract, and optic nerves.
Table 1. List of anatomical nomenclature
Nomenclature used in sea lion atlas
Nomenclature reported in Nomina Anatomica Veterinaria (fifth edition).
The atlases of the brain and hippocampus of the California sea lion were constructed from native (i.e., no image processing) transverse and oblique T2-weighted images, respectively. These images were created using eFilm Lite 2.1.2 (Merge Healthcare, Milwaukee, WI) from the Digital Imaging and Communication in Medicine (DICOM) files saved during the TSE sequences. T2-weighted images were used in the label schematics because these images are very sensitive to minute changes in water concentration and are therefore useful in illustrating pathology within the brain. The oblique T2-weighted images were selected for the hippocampal atlas because these images were exceptional in defining hippocampal anatomy, as previously described (Goldstein et al., 2008).
Volume Analysis of Brain Structures
Visualization of MR images was completed first on the MRI unit. Post-processing, segmentation, 3D reconstructions, and volume analysis were also performed using the software program AMIRA 4.1.1 (Mercury Computer Systems, San Diego, CA). 3D reconstructions and volumes of brain structures that were determined included whole brain; cerebrospinal fluid (CSF) of the total, left, and right brain; CSF of the total, left, and right cerebral ventricles; GM and WM of the entire brain; GM and WM of the total, left, and right cerebral hemispheres; and GM and WM of the cerebellum including the brainstem. Volumes of the left and right hippocampus and associated structures that were determined included lateral ventricle (ventral horn); hippocampal sulcus; hippocampus and parahippocampal gyrus combined; hippocampus alone; parahippocampal gyrus alone; and parahippocampal gyrus WM and GM. The WM of the parahippocampal gyrus most likely contained fibers of the subiculum because of the inability to identify boundaries between WM of the parahippocampal gyrus and WM of the subiculum. Segmenting brain structures, creating 3D reconstructions, and volume analysis used methods similar to those described by Montie et al. (2008) with modifications explained below.
Brain CSF, GM, and WM.
To determine the volumes of brain structures, a brain surface mask created from the native T2-weighted transverse images was produced to determine edges for digital removal of nearby blubber, muscle, skull, and other head structures. The mask was constructed by manually tracing the surface of the brain and deleting all pixels outside this trace for each MR image. The whole brain volume was calculated by integrating the area of the selected tissue for each slice of the brain surface mask. The caudal boundary of the brain was defined by the caudal aspect of the foramen magnum. Virtual brain weight was calculated by multiplying the total brain segmented volume by the assumed specific gravity of brain tissue, 1.036 g/cm3 (Stephan et al., 1981).
Total brain CSF volumes were determined by threshold segmentation of the brain surface mask (obtained from the native T2-weighted transverse images) followed by manual editing of each slice. Specifically, this procedure involved thresholding for signal intensity ranges that captured the boundaries of brain CSF followed by visual inspection and manual editing to ensure that CSF was properly defined. From this segmentation, pixels representing CSF of the left and right hemispheres were manually selected and defined as a new label field (i.e., a set of images where pixels are selected to represent different anatomical structures). Pixels of the left and right ventricular system that captured the lateral, third, and fourth ventricles were manually selected and defined as another label field. The volumes of these structures (i.e., total CSF, left and right CSF, total ventricles, left and right ventricles) were determined three separate times. The mean volume (cm3) and standard deviation of these structures were reported.
Total GM and WM volumes were determined by a combination of manual and threshold segmentation of processed, transverse PD-weighted images. Processing involved deleting the pixels outside the brain surface mask and deleting the pixels of each CSF label field from the native PD-weighted transverse images. From the processed PD-weighted images, the segmentation procedure involved thresholding for signal intensity ranges that captured the boundaries of GM and WM followed by visual inspection and manual editing to ensure that GM and WM were properly defined. The volume of total GM and WM was then determined. From this segmentation, pixels representing cerebral GM and WM and pixels representing cerebellum and brainstem GM and WM were manually selected and defined as a new label field. The volumes of total cerebral GM and WM and cerebellum plus brainstem GM and WM were determined. The pixels defining GM and WM of the cerebellum were combined with the pixels representing GM and WM of the brainstem because the boundaries of the cerebellum and brainstem were not easily recognized. Pixels defining the left and right cerebral GM and WM were manually selected and the volumes of these structures determined. For each structure, volumes were determined three times. The mean volume (cm3) and standard deviation of these structures were reported.
Volumes of the left and right hippocampus and associated structures were determined by manual segmentation of native, oblique T2-weighted images. The T2-weighted images were used because they were better at highlighting fluid structures surrounding the hippocampus compared with the PD images. These fluid structures served as boundaries of the hippocampus and were defined by higher signal intensities. The atlas by Dua-Sharma et al. (1970) and the hippocampus atlas we constructed served as guides for segmenting the left and right hippocampus and associated structures. For these volume calculations, the various structures of the hippocampal formation (i.e., subiculum, cornu ammonis, and the dentate gyrus) could not be adequately distinguished. These brain regions and the fimbria and alveus were collectively grouped and referred to as the hippocampus. Structure volumes were determined three times, separately. The mean volume (mm3) and standard deviation of these structures were reported.
The percentage of brain occupied by the left or right hippocampal structure was calculated by dividing that structure's volume (i.e., from the native oblique T2-weighted images) by the sum of the WM and GM volumes of the whole brain (i.e., from the processed PD-weighted images) multiplied by 100. The percentage of cerebral hemisphere occupied by the left or right hippocampal structure was calculated by dividing the structure volume (i.e., from native T2 weighted images) by the sum of the left or right cerebral WM and GM volumes (i.e., from processed PD weighted images) multiplied by 100.
RESULTS AND DISCUSSION
3D Reconstruction of the Brain Surface from Magnetic Resonance Images
Comparing the 3D reconstruction of the California sea lion brain to photographs of formalin-fixed brains of other mammals (i.e., from the Comparative Mammalian Brain Collection website, http://www.brainmuseum.org/index.html) showed the resemblance of the sea lion brain to the brains of other species belonging to the Order Carnivora, specifically carnivores within the Suborder Caniformia (Fig. 1). The California sea lion brain was similar in shape to representative species in the Family Canidae [e.g., the domestic beagle (Canis familiaris), the wolf (Canis lupus), and the coyote (Canis latrans)]; the Family Ursidae [e.g., the polar bear (Ursus maritimus)]; and the Family Mustelidae [e.g., the American mink (Neovison vison) and the American badger (Taxidea taxus)]. Compared with the canids, ursids, and mustelids, the California sea lion brain was expanded laterally, with large frontal, temporal, and occipital lobes. Qualitative comparisons of the degree of neocortical folding (or gyrification) among these carnivores indicated that the brain of the California sea lion contained more secondary folds and sulci than the American mink, the American badger, the domestic beagle, the wolf, the coyote, and the polar bear. In addition, the pattern of folds and fissures in the sea lion brain was very different from these carnivores. For example, the Sylvian fissure in the sea lion brain was perpendicular to the ventral surface of the brain, whereas in the domestic dog, the sulcus was obtuse (∼135°) to the ventral surface. The Sylvian fissure in the sea lion also extended deeper into the brain towards the longitudinal cerebral fissure compared with the shallower fissures in the domestic dog.
The increase in gyri and sulci in the California sea lion compared with canids and mustelids may be best explained by the larger brain size in the sea lion. Neocortical folding has been shown to correlate with brain size in large-brained mammals in many different lineages, as reviewed by Striedter (2005). It has been speculated that the neocortex tends to fold in large brains because the most efficient way to increase the area of the neocortex without dramatically increasing neocortex thickness is for the neocortex to fold inward [as reviewed by Striedter (2005)]. If the neocortex were to have ballooned outward, the expansion without this folding would have yielded very large heads and long intracortical connections, both unfavorable factors during natural selection [as reviewed by Striedter (2005)].
Neuroanatomical Atlas from Magnetic Resonance Images
Figures 2–21 display a rostral-to-caudal sequence of T2-weighted, 2.5 mm-thick transverse MR brain sections at 5 mm intervals. Panels A display a sagittal section showing the orientation and level at which the T2 section was taken; Panels B illustrate the position of the brain in the transverse plane relative to surrounding head structures of the T2-weighted image; panels C show labeled versions of each brain section. The right side of the images corresponds to the left side of the brain, which is the traditional method in showing radiological images. These figures demonstrate undisturbed spatial relationships among brain structures and surrounding head anatomy obtainable by MR imaging of live animals.
Figures 22–35 display a ventral-to-dorsal sequence of T2-weighted, 2.5-mm thick oblique MR brain sections at 2.5 mm intervals of the hippocampal region. Panels A display a sagittal section showing the orientation and level at which the T2 section was taken; panels B illustrate the position of the brain in the oblique plane relative to surrounding head structures of the T2-weighted image; panels C show labeled versions of each brain section. The right side of the images corresponds to the left side of the brain. The images obtained in the oblique plane provided a better view of the hippocampus compared with the images obtained in the transverse plane.
The MR images illustrate distinguishing features of the California sea lion telencephalon. The size of the olfactory bulb relative to the brain is small (Figs. 2C–3C); comparisons indicated that the bulb in canids, ursids, and mustelids are larger. The decrease in size of this brain structure in California sea lions may be related to the possibility that sea lions rely less on olfactory signals to detect prey and predators than terrestrial carnivores. It is very possible that pinnipeds do not have active olfaction underwater (Denhardt, 2002).
The neocortex was highly convoluted and contained many secondary gyri and sulci that were not identifiable using the domestic dog atlas (Figs. 2C–21C). The suprasylvian gyrus was expanded laterally (Figs. 5C, 6C), more so than the gyrus of the domestic dog. Despite the large hemispheres, the corpus callosum appears to be small (Figs. 8C–13C).
Structures of the basal ganglia (i.e., the caudate nucleus, the putamen, and the globus pallidus) were identified (Figs. 7C–12C). Interestingly, the caudate nucleus in the California sea lion seems to have a very small tail. The putamen was difficult to recognize and seemed to be mixed with fiber bundles.
The amygdala was evident in the rostral portion of the temporal lobes (Figs. 10C, 22C), whereas the hippocampus was located near the center of the sections in the medial wall of the temporal lobes (Figs. 11C–13C, 23C–33C). The boundaries of the hippocampus were best observed in native T2-weighted images rather than PD-weighted images. This finding can be best explained by the CSF surrounding the hippocampus, as observed by the hyperintensity of the ventral horns of the lateral ventricles (lateral and dorsal borders) (Figs. 11C–13C, 23C–33C) and the hyperintensity of the hippocampal sulcus (medial border) (Figs. 11C–12C, 22C–29C). The structures of the hippocampal formation were best visualized in the oblique T2-weighted images (Figs. 23C–33C). These structures included the cornu ammonis or hippocampus proper (Figs. 23C–33C), the subiculum (Figs. 24C–29C), and the dentate gyrus (Figs. 26C–30C). The small size of the dentate gyrus made its identification very difficult in some of the sections. WM tracts of the hippocampus, the alveus and fimbria, were also identified (Figs. 24C–32C). The parahippocampal gyrus was easily recognized in the MR images (Figs. 11C–13C, 23C–33C).
One very interesting finding was that the California sea lion hippocampus was found mostly in the ventral position with very little extension dorsally (Figs. 11C–13C, 23C–33C). However, in canids and mustelids, the hippocampus is present in the ventral position but also extends dorsally above the thalamus (Dua-Sharma et al., 1970; Welker et al., 2009).
The MR images revealed a large diencephalon in the California sea lion. The hypothalamus (Figs. 10C–11C) and thalamus (Figs. 10C–14C) were easily recognized in the MR images. The optic nerves (Figs. 5C–8C), optic chiasm (Fig. 9C), and pituitary gland (Figs. 10C–12C) were also observed. The mammillary body was visualized, protruding downward towards the pituitary gland (Fig. 11C). The pineal gland was very prominent, located in the midline above the caudal commissure (Fig. 13C). This finding was expected, as the pineal gland is known to be exceptionally large in pinnipeds (Cuello and Tramezzin, 1969; Turner, 1888). The habenular nucleus was also located just underneath the pineal gland (one on each side) (Fig. 12C).
The rostral colliculus (i.e., superior colliculus in bipeds) was easily identified, as was the caudal colliculus (i.e., inferior colliculus in bipeds) (Fig. 14B). Neither of these structures appeared to be particularly enlarged.
The MR images showed typical characteristics of the carnivore metencephalon and myelencephalon. Auditory pathways were observed, including the cochlea (Fig. 15C), the vestibulocochlear nerve (Fig. 15C), and the lateral lemniscus (Fig. 14C). The trigeminal ganglion (Figs. 10C–12C) and trigeminal nerve (Figs. 13C–15C) were identified, as well as the trigeminal nucleus (Fig. 17C). The cerebellum was large, and GM and WM were easily distinguishable (Figs. 15C–21C). Hindbrain structures including the pons (Figs. 13C–15C), reticular formation (Figs. 13C–20C), pyramidal tract (Figs. 16C–17C, 20C–21C), and medulla oblongata were identified (Figs. 20C, 21C).
CSF and Cerebral Ventricles.
The three major processes of the lateral ventricles were easily recognized including the rostral or frontal horn (Figs. 7C–14C), the caudal or occipital horn (Fig. 15C), and the ventral or temporal horn (Figs. 11C–14C). The third ventricle appeared as a thin slit between the two thalami or hypothalami (Figs. 10C–12C). The interventricular foramen (i.e., a fluid connection where CSF flows from the lateral ventricles to the third ventricle) was observed (Fig. 11C). The mesencephalic aqueduct (i.e., a thin, fluid connection where CSF flows from the third to the fourth ventricle) was identified but characterized by a signal void (Figs. 13C–14C). The lack of CSF signal in the mesencephalic aqueduct represents a flow artifact secondary to pulsatile CSF flow (Feinberg and Mark, 1987; Malko et al., 1988; Lisanti et al., 2007). The fourth ventricle was visualized as a tent-like structure just dorsal of the reticular formation and ended at the obex (Figs. 16C–19C). The lateral recess of the fourth ventricle (i.e., a site where CSF leaves the ventricular system and enters the subarachnoid space) was also observed (Fig. 18C). The choroid plexus (i.e., vascular tufts responsible for CSF production) in the lateral ventricles (Figs. 10C–12C) and the fourth ventricle (Fig. 18C) were recognized. Choroid plexus appears to be present in the third ventricle as well (Fig. 12C).
Volumes of Brain Structures
Brain CSF, GM, and WM.
Segmentation of the transverse T2-weighted images was used to delineate the brain surface and calculate whole brain volume (Table 2). The volume of the entire brain was 301.71 cm3, which included CSF of the subarachnoid space and cerebral ventricles. The virtual brain weight (calculated by multiplying the measured volume by the specific gravity of brain tissue) was 312.6 g. This estimate was very similar to the actual brain weight measured after fixation (i.e., 306 g).
Table 2. Brain volumes of the California sea lion
Total cerebrospinal fluid
38.96 ± 2.08
Left cerebrospinal fluid
19.94 ± 1.06
Right cerebrospinal fluid
19.02 ± 1.03
7.59 ± 0.12
3.76 ± 0.05
3.83 ± 0.07
Total gray matter
164.34 ± 6.57
Total white matter
80.07 ± 2.62
Total cerebral gray matter
129.39 ± 5.22
Left cerebral gray matter
64.82 ± 2.84
Right cerebral gray matter
64.06 ± 2.12
Total cerebral white matter
54.42 ± 1.87
Left cerebral white matter
27.00 ± 1.12
Right cerebral white matter
27.43 ± 0.89
Cerebellum and brainstem gray matter
25.65 ± 1.13
Cerebellum and brainstem white matter
34.94 ± 1.45
The volumes of CSF and cerebral ventricles were determined from segmentations of the transverse T2-weighted images, after digital removal of nearby blubber, muscle, skull, and other head anatomy (Table 2). The CSF volumes of the left and right hemispheres were approximately equal (Table 2). In addition, the CSF volumes in the cerebral ventricles of the left and right hemispheres were similar (Table 2).
The volumes of GM and WM of the brain were estimated from segmentations of the transverse PD-weighted images, after digital removal of nearby blubber, muscle, skull, and other head anatomy and removal of total CSF (Table 2). The volumes of either the GM or the WM of the left and right cerebral hemispheres were approximately equal (Table 2). 3D reconstructions of GM and CSF of the subarachnoid space of the left and right hemispheres were constructed (Fig. 36A). The GM and subarachnoid CSF in the 3D reconstruction were then removed to reveal the underlying WM of the left and right hemispheres (Fig. 36B).
The volumes of the ventral horn of the lateral ventricles, the hippocampal sulcus, the hippocampus (including alveus and fimbria, dentate gyrus, cornu ammonis, and subiculum), and the parahippocampal gyrus were determined from segmentations of oblique T2-weighted images (Table 3). The volumes of the left and right hippocampus were approximately equal (Table 3). In addition, the left and right parahippocampal gyri were approximately equal in volume (Table 3). Both the volumes of the left and right hippocampus were 0.84% of the volumes of the left and right cerebral hemispheres (Table 3). Both the volumes of the left and right parahippocampal gyrus were approximately 1.6% of the volumes of the left and right cerebral hemispheres (Table 3). 3D reconstructions of the cerebral ventricles and their association with the hippocampi were created (Fig. 36C). The ventricles were then removed to reveal the underlying hippocampi, fornix, septal nucleus, and mammillary bodies (Fig. 36D). The 3D reconstruction of the California sea lion hippocampus revealed its ventral position with very little extension dorsally (Fig. 36D), as previously mentioned. In the sea lion, only the fornix was found above the thalamus. This finding in the sea lion is different from the domestic dog and the American mink (see the Comparative Mammalian Brain Collection by Welker et al. (2009), http://www.brainmuseum.org/); in these terrestrial relatives, the hippocampus is present in the ventral position but also extends dorsally above the thalamus.
Table 3. Volumes of the hippocampus and associated structures in the California sea lion
Percentage of total brain occupied by the left or right structure = left or right structure volume (from native T2-weighted oblique images) divided by the white matter (WM) plus gray matter (GM) volumes of the whole brain (from processed PD-weighted images) multiplied by 100%.
Percentage of left or right cerebral hemisphere occupied by the respective left or right structure = left or right structure volume (from native T2-weighted oblique images) divided by the respective left or right cerebral WM plus left or right cerebral GM volumes (from processed PD-weighted images) multiplied by 100%.
The WM of the parahippocampal gyrus may contain subiculum WM.
This article presents the first anatomically labeled MRI-based atlas of a pinniped brain. It is different from previous MRI-based atlases of marine mammals in that it was created from imaging a live animal. This study also presents a quantitative approach to determine the size of brain structures, such as the hippocampus, from MR images of live California sea lions.
Live MRI scanning coupled with volumetric analysis can be used not only as a tool to study brain evolution in pinnipeds but also to investigate the impacts of biological, chemical, and physical agents on marine mammal health (Montie, 2006). Of particular concern are the acute, chronic, and possible developmental effects of DA on the brain of California sea lions (Silvagni et al., 2005; Goldstein et al., 2008; as reviewed by Ramsdell and Zabka, 2008). The MRI atlas presented here of a neurologically normal California sea lion will allow us to better evaluate the hippocampus in the oblique plane and determine the volumes of brain structures. This approach will help in deciphering the acute, chronic, and developmental effects of DA exposure, or exposure to other pollutants, in wild California sea lions, other marine mammals, and wildlife in general.
The authors thank all the staff and volunteers at the Marine Mammal Center, especially Elizabeth Wheeler for performing the necropsy of this animal. They also thank Dr. Cheryl Cross for providing gross images of this California sea lion brain, and Dr. Heather Harris and Dr. Felicia Nutter for their veterinary assistance. They thank the University of Wisconsin and Michigan State Comparative Mammalian Brain Collection, and the National Museum of Health and Medicine for providing gross and histological images of carnivore brains. The preparations of those images were funded by the National Science Foundation, as well as the National Institutes of Health.