Volumetric Analysis of the African Elephant Ventricular System


  • Busisiwe C. Maskeo,

    1. School of Anatomical Science, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, Republic of South Africa
    Search for more papers by this author
  • Muhammed A. Spocter,

    1. School of Anatomical Science, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, Republic of South Africa
    2. Department of Anthropology, The George Washington University, Washington, DC
    Search for more papers by this author
  • Mark Haagensen,

    1. Department of Radiology, University of the Witwatersrand-Donald Gordon Medical Centre, Johannesburg, Republic of South Africa
    Search for more papers by this author
  • Paul R. Manger

    Corresponding author
    1. School of Anatomical Science, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, Republic of South Africa
    • School of Anatomical Sciences, Faculty of Health Sciences, University of the Witwatersrand, 7 York Road, Parktown, 2193, Johannesburg, Republic of South Africa.
    Search for more papers by this author


This study used magnetic resonance imaging (MRI) to determine the volume of the ventricular system in the brain of three adult male African elephants (Loxodonta africana). The ventricular system of the elephant has a volume of ∼240 mL, an order of magnitude larger than that seen in the adult human. Despite this large size, allometric analysis indicates that the volume of the ventricles in the elephant is what one would expect for a mammal with an ∼5 kg brain. Interestingly, our comparison with other mammals revealed that primates appear to have small relative ventricular volumes, and that megachiropterans and microchiropterans follow different scaling rules when comparing ventricular volume to brain mass indicating separate phylogenetic histories. The current study provides context for one aspect of the elephant brain in the broader picture of mammalian brain evolution. Anat Rec, 2011. © 2011 Wiley-Liss, Inc.


A comprehensive description of the structure of the ventricular system in elephants is provided by Shoshani et al. (2006), who comment that for the most part, except for absolute volume and proportions, they resemble the structures seen in the human brain and many other mammals. Shoshani et al. (2006) detail that the lateral ventricles are broad and flat, but have foreshortened frontal horns, small to absent occipital horns, and vertically oriented temporal horns. The olfactory ventricles are large and connect to the lateral ventricles through an inferior olfactory recess leading from the ventral surface of the frontal horn. Shoshani et al. (2006) further report that the third ventricle is interrupted by a large interthalamic adhesion, that there appears to be a significant dorsal cerebellar recess of the fourth ventricle and choroid plexus seen in the lateral, third and fourth ventricles. It is noted by Shoshani et al. (2006) that none of the ventricles of the elephant show any specific evidence of hydrocephaly.

In the work of Cheng et al. (2010), it is stated that the structure of the ventricles of the brain play a vital role in the flow of cerebrospinal fluid (CSF), in that the CSF passes in a pulsatile manner that is related to the pulsation of the heart. The volume of these ventricles is vital in indicating the amount of CSF that is located within them (Shoshani et al., 2006) and understanding the amount of CSF contained is important as an indicator of the physiology of the brain in the species concerned. Ventricular volume may reveal certain requirements of the brain since the CSF has a variety of functions including shock absorption, distribution of neurotransmitters and nutrients, removal of toxins and CSF protects the brain during blood pressure fluctuations (Di Terlizzi and Platt, 2006).

Systematic data regarding the total ventricular volume (TVV) is available for a range of mammalian species including primates and insectivores (Stephan et al., 1981), and microchiroptera and megachiroptera (Baron et al., 1996). Given the likelihood that there are significant correlations between brain mass and TVV (Finlay and Darlington, 1995), the current study aimed to analyse this previously published data in relation to newly calculated TVVs for African elephants. Although elephant ventricles are indeed absolutely large (Shoshani et al., 2006), the question of whether they are relatively large remains open. This question follows on from the reports regarding the relatively large size of the lateral ventricles in the closely related manatees (Welker, 1990). One possibility is that the Afrotherians in general have a common phylogenetic enlargement of the ventricular system in comparison to other mammals. The second possibility is that the ventricular volume in elephants is what would be expected for a mammal with its brain mass, and thus the qualitative enlargement reported for the manatee ventricles would be an adaptive response to their current or past environment. Thus, using magnetic resonance imaging (MRI), the TVVs of three male African elephants was calculated and compared with that previously reported for other mammals.


In this study, data on brain mass and TVV was obtained from two sources, previously published literature and analyses based on magnetic resonance (MR) images undertaken specifically on three African elephant brains (LA1, LA2, and LA3, Manger et al., 2009, all males between the ages of 20 and 30 years). Data on brain mass and the TVV of microchiropterans (n = 222 species) and megachiropterans (n = 47 species) were taken from Baron et al. (1996), whereas data for insectivores (n = 28 species) and primates (n = 46 species) were taken from Stephan et al. (1981).


Brains from the three African elephants (Loxodonta africana, LA1, LA2, and LA3) underwent MRI to obtain measurements of TVVs. The brains were fixed and stored as described in Manger et al. (2009), the brains being perfused in situ and removed from the skull immediately and following postfixation were scanned in coronal, sagittal, and horizontal planes. The specimens were scanned on a Philips 1.5 Tesla Intera System (Eindhoven, The Netherlands), using all three elements of the head and neck coil. The brains were removed from their storage containers, drained of excess fluid and placed in the head coil wrapped in a dry sheet, thus being exposed directly to air, which also partly entered the ventricles. After testing different scan parameters, the following sequence was selected as giving the best detail and the least artefact (especially at air-fluid interfaces). The selected T1 weighted inversion recovery sequence consisting of 2-mm slices without gap, had a time to repeat of between 6.5 and 10.9 s depending on the number of slices, a time to echo of 10 ms and a time to invert of 300 ms. The number of signal averages varied between 3 and 4 with a flip angle of 90° and an echo train of 10. The scan times varied between 15 and 25 min. The antifreeze liquid in which the brains were stored showed high signal on both T1 and T2 weighted sequences and the routine clinical T1 and T2 sequences produced very similar T2 like images of the brain specimens. This is possibly related to the lack of water in the tissues of the specimen secondary to the fixation and storage process. The images were processed using the freely available open source software program OsiriX (Rosset et al., 2004; www.osirix-viewer.com).

Calculating Ventricular Volume from MR Images

The calculation of TVV in African elephants has, to our knowledge, not been undertaken previously, and herein includes the lateral ventricles, the third ventricle, the cerebral aqueduct and the fourth ventricle. For each brain, the ventricles were measured for volume by tracing around the ventricular walls in each of the slices showing the presence of any ventricle (LA1 – 103 slices, LA2 – 98, and LA3 – 99). This was done in the coronal plane, which highlighted the ventricular borders most clearly. It is important to note here that the olfactory bulbs could not be dissected properly from the elephant cranium as part of the whole brain (Manger et al., 2009), therefore they were not available to include in total ventricle measurement. The OsiriX program was used for the tracing and measurement of the area of the ventricles in each slide, the sum of which multiplied by slice thickness (2 mm) and converted to volume measurements provided the TVV. Figure 1 shows the appearance of the ventricles and their borders at different coronal levels in MR images.

Figure 1.

MR images showing the elephant ventricular system at different levels, in the coronal plane. A is the most rostral section, H the most caudal. AE show the lateral ventricles; D shows the large interthalamic adhesion in the third ventricle; F-H show the fourth ventricle with the large cerebellar recess in G. Scale bar = 5 cm.

Statistical Analyses

The data was logarithmically transformed to the base 10 for allometric analysis. We then divided the data into five groups, these being elephants, primates, megachiropterans, insectivores, and microchiropterans and undertook three separate analyses of the transformed data. First, standardized major axis, or reduced major axis (see Warton et al., 2006 for discussion of the terminology), was used to describe and compare the bivariate relationship of the different groups. The software SMATR ver. 2.0 (Warton et al., 2006; www.bio.mq.edu.au/ecology/SMATR) was used to test for common slopes between the groups, and where present, test for shifts in elevation or along the common axis. Second, ordinary least squares regressions were calculated for the primate, megachiroptera, insectivore and microchiropteran groups. The probability of using these regressions to correctly predict TVV for individual observations in the elephant was assessed (Sokal and Rohlf, 1995, p. 469), and prediction intervals drawn to aid description.


The Ventricular System of the African Elephant

The current MRI observations confirmed the anatomy of the elephant ventricular system as described in detail by Shoshani et al. (2006). The basic ventricular formation is typical of mammals, with the ventricles containing choroid plexus in the expected regions (Fig. 1). As described by Shoshani et al. (2006), the frontal and occipital horns of the lateral ventricles are foreshortened, and the olfactory ventricle is connected to the anteroventral portion of the lateral ventricle through a small opening (Fig. 1B). The third ventricle is dominated by a large interthalamic adhesion (Fig. 1D) and the fourth ventricle has a significant dorsal expansion, forming a large cerebellar recess (Fig. 1G). In terms of TVV, the results obtained for the three elephants were: LA1 – TVV = 223.58 mL, brain mass = 5145 g; LA2 – TVV = 249.06 mL, brain mass = 5250 g; LA3 – TVV = 243.68 mL, brain mass = 4835 g.

Brain Mass Versus TVV in Primates, Megachiroptera, Microchiroptera and Insectivores

When the brain masses were compared with the total volume of the ventricles in primates, megachiropterans, microchiropterans, and insectivores, in all cases a strongly statistically significant relationship was observed (Fig. 2); however, a test for commonality of slopes indicated that each group was significantly different from the others (Fig. 2). For primates, a slightly negative allometry was observed (slope = 0.975, r2 = 0.929, P = 6.0 × 10−27); for megachiropterans a positive allometry was observed (slope = 1.229, r2 = 0.965, P = 2.7 × 10−34); microchiropterans showed a positive allometry (slope = 1.104, r2 = 0.829, P = 2.3 × 10−86), whereas insectivores showed the greatest positive allometry (slope = 1.469, r2 = 0.935, P = 5.9 × 10−17).

Figure 2.

Graphical plots depicting the relationships between brain mass and TVV in elephants, primates, insectivores, microchiropterans, and megachiropterans. The large upper graph shows the raw data on log transformed scales with the associated statistical equations describing the lines of best fit. The lower six smaller graphs represent the relationship of the elephant data to the 95% confidence intervals for the different groups studied on log transformed data.

Elephant TVV Compared with Other Mammals

To compare the relative size of the ventricular volume in elephants to the other mammals, we calculated the 95% confidence intervals for each group (these are presented graphically for ease of interpretation in Fig. 2). When the elephant data was compared with the 95% confidence intervals obtained for megachiropterans, microchiropterans, and insectivores, the elephants clearly fell well within these intervals for all three groups, although on the lower side of the range when compared with insectivores. This indicates that the relative volume of the ventricles in elephants is not statistically significantly different from these three mammalian groups. When the elephants are compared with primates, it was seen that the elephants fall outside and above the 95% confidence interval for this group, indicating that they have statistically significantly larger relative TVVs than would be predicted from ventricular volumes of primate brains. These comparisons were taken further by grouping all the mammals (primates, megachiropterans, microchiropterans, and insectivores), and in this instance, the elephants had larger relative ventricular volumes than would be predicted, as they fell above the 95% confidence interval. Because the insectivores exhibited the steepest positive allometry, a comparison of the mammalian volumes minus the insectivores (primates, megachiropterans, and microchiropterans) was also tested against the elephants, and again the elephants showed larger relative ventricular volumes, falling above the 95% confidence interval for this grouping. These last two comparisons however, appear to be affected by the inclusion of primates, and indicate that the inclusion of the primates tends to skew the comparison in such a way as to make the elephants appear to have significantly large ventricles. It would seem more prudent to conclude from these comparisons that the primates in fact have smaller relative ventricular volumes in comparison to other mammals.


It was found that the relationship between brain mass and TVV is predictable across mammals, but that this predictability is specific to each mammalian order. The absolute volume of the elephant ventricles is the largest recorded to date, but some larger brained cetaceans, which remain to be measured, may exceed that observed for the elephant. In relative terms, TVV observed for the elephant is what would be expected for a mammal with an ∼5 kg brain. Interestingly, primates appear to have small relative TVVs in comparison to other mammals.

Order Specific Predictability of TVVs Across Mammalian Species

According to Finlay and Darlington (1995), there is predictability, or linked regularities, between brain mass and the structures that make up that mass. For example, they demonstrate that several major subdivisions of the actual matter of the brain covary in specific proportions in relation to the overall mass of the brain. Although certain exceptions to this predictability have been noted (e.g., Kaas and Collins, 2001), as a general rule it is applicable across mammalian species as a whole. Finlay and Darlington (1995) and their subsequent studies (e.g., Finlay et al., 2001; Reep et al., 2007; Yopak et al., 2010) have not examined that part of the brain, the ventricles, which do not contribute to total brain mass. In this study, we found that the TVV, whereas strongly correlated with brain mass, showed this relationship far more strongly when a specific order was examined. Moreover, there were statistically significant differences between the orders, in agreement with the concept of predictability of brain structures within orders, but differences between orders (Manger, 2005). Interestingly, in the current study we analyzed megachiropterans and microchiropterans as two different groups, being considered two distinct orders for the analysis, and there is clear separation between these two groups, but strong cohesion within these groups. This distinction between the megachiropterans and microchiropterans adds yet another neural trait indicating that these two groups should be regarded as distinct orders of mammals with different evolutionary trajectories (Pettigrew, 1986; Pettigrew et al., 1989, 2008; Maseko and Manger, 2007; Maseko et al., 2007; Kruger et al., 2010; Dell et al., 2010).

Absolute and Relative TVV of the Elephant

The TVV of the elephants determined in the current study is the largest measured to date. At ∼240 mL, the volume of the ventricular system of the elephant is an order of magnitude larger than that observed for the human brain, which had the second largest ventricular volume (∼20 mL, Stephan et al., 1981) and brain mass in the dataset used in the current analysis. The volume and appearance of the ventricles of the three male African elephants quantified in this study appear similar to that described for the seven elephants (three female Asian, three female African, and 1 male African, although this male was newborn) by Shoshani et al. (2006, see their Fig. 6), indicating no potential problems with fixation and subsequent processing of the specimens analyzed in the current study. In contrast to this, is the size of the ventricles in a single female African elephant (brain mass less than 4 kg) that appear to be very small (Hakeem et al., 2005). The appearance of small ventricles in this particular specimen likely results from an extended period of postfixation ex situ (the images of the brain appear to show significant distortion and the top of the brain appears flat compared with the roundness of the dorsal surface of the brain in our specimens) resulting in collapsing and shrinking of the ventricles—for example, no significant cerebellar recess, as noted in the current study and by Shoshani et al. (2006), is evident in the Hakeem et al. (2005) specimen. It would be of interest to examine other species with brains larger than humans, such as the larger-bodied cetaceans, that may have larger TVVs than seen in the elephant (although in general the cetaceans do not appear to have a large ventricular system, so the elephant may actually have the largest).

On the other hand, in relative terms, the TVV of the elephant appears to be a reflection, or consequence, of having a 5-kg brain. When compared with the insectivores, microchiropterans, and megachiropterans, the elephants have a ventricular volume that would be expected for such a large brain. It is only when compared with primates (see below) that the elephants appear to have relatively large ventricles. Given this consistency in size with three of the four groups analyzed, the size, either absolute or relative, of the elephant ventricular system cannot be thought of as adaptive in nature. Subsequently, no functional correlates above and beyond that normally associated with the mammalian ventricular system can be ascribed to the ventricular system of the elephant.

Primates Have Relatively Small TVVs

One of the surprising results to emerge in this study was that in comparison to the other mammals investigated, primates were found to have relatively small ventricles. One potential confound with this finding is that in terms of the overlap of brain masses and ventricular volumes of primates (brain masses 1.8–1330 g, TVV 0.01–18.752 mL) with the insectivores (brain masses 0.11–6.1 g, TVV 0.005–0.097 mL), microchiropterans (brain masses 0.08–2.6 g, TVV 0.0004–0.023 mL), and megachiropterans (brain masses 0.5–9 g, TVV 0.004–0.201 mL) was limited to the lower end of the range of primate brains used in this analysis. This difference may lead to a potentially false positive result and indicates the need to obtain data from several larger brained mammalian species such as the carnivores, artiodactyls, and cetaceans to determine whether this is a real phenomenon, or a result of unavoidable sampling bias in the analysis. If, however, the current finding was to be confirmed with further analysis, this would present an avenue of interest to investigate further in terms of primate brain evolution, development and adaptation in comparison to other mammals.


We would like to thank Dr. Theo Nel and staff at Wits-DGMC, for the kind use of the MR scanner and their help during this project, Dr. Hilary Madzikanda of the Zimbabwe Parks and Wildlife Management Authority, and Dr. Bruce Fivaz and the team at the Malilangwe Trust, Zimbabwe.