The Balance Hypothesis for the Avian Lumbosacral Organ and an Exploration of Its Morphological Variation

Abstract

Birds (Aves) exhibit exceptional and diverse locomotor behaviors, including the exquisite ability to balance on two feet. How birds so precisely control their movements may be partly explained by a set of intriguing modifications in their lower spine. These modifications are collectively known as the lumbosacral organ (LSO) and are found in the fused lumbosacral vertebrae called the synsacrum. They include a set of transverse canal-like recesses in the synsacrum that align with lateral lobes of the spinal cord, as well as a dorsal groove in the spinal cord that houses an egg-shaped glycogen body. Based on compelling but primarily observational data, the most recent functional hypotheses for the LSO consider it to be a secondary balance organ, in which the transverse canals are analogous to the semicircular canals of the inner ear. If correct, this hypothesis would reshape our understanding of avian locomotion, yet the LSO has been largely overlooked in the recent literature. Here, we review the current evidence for this hypothesis and then explore a possible relationship between the LSO and balance-intensive locomotor ecologies. Our comparative morphological dataset consists of micro-computed tomography (μ-CT) scans of synsacra from ecologically diverse species. We find that birds that perch tend to have more prominent transverse canals, suggesting that the LSO is useful for balance-intensive behaviors. We then identify the crucial outstanding questions about LSO structure and function. The LSO may be a key innovation that allows independent but coordinated motion of the head and the body, and a full understanding of its function and evolution will require multiple interdisciplinary research efforts.

Fig. 1

Birds are hypothesized to have two sets of balance organs, one in the inner ear (A) (common to terrestrial vertebrates) and a second within the synsacrum (B), termed the lumbosacral organ (LSO). The LSO is housed within an expanded vertebral canal, which exhibits a set of canal-like recesses (lumbosacral transverse canals [LSTCs]; C).

Fig. 2

A contrast-enhanced CT scan demonstrates the soft tissue morphology of the avian LSO and the space within the LSTCs. A parasagittal section (A) shows the LSTCs and the expansion in the vertebral canal for the glycogen body. A corresponding lateral view of a 3D model from the CT scan demonstrates the accessory lobes (B). A horizontal section (C) and the corresponding dorsal view of the soft tissue in the model (D) show the glycogen body within the dorsal groove of the spinal cord. Ridges evident in a dorsal view of the endocast of the vertebral canal (E) result from the LSTCs. A transverse section of the third canal from the anterior end (F) is paired with labeled soft tissue elements (G).

Fig. 3

The measure of LSTC prominence is calculated from a series of cross-sectional areas from the vertebral canal. The diagram of the parasagittal section (A) illustrates how we selected cross-sections approximately perpendicular to the long axis of the vertebral canal. The labeled sections are examples corresponding with the formula for the dimensionless ratio calculation provided in the text. The transverse section diagrams (B) illustrate the cross-sectional areas of the vertebral canal for canal and intercanal sections. A plot of the cross-sectional areas is provided in (C), with a dashed line representing the running mean. A plot of the corresponding ratios is in (D). The LSTC prominence metric is defined as the standard deviation of these ratios.

Fig. 4

The morphology of the synsacral vertebral canal varies across bird species. These are sagittal and parasagittal sections from μ-CT scans of the synsacra of three bird species paired with lateral and dorsal views of the corresponding vertebral canal endocasts. Although A. pacificus and S. humboldti have similar values for the metrics we calculated, the shape of their endocasts differ, particularly in the orientation of the LSTCs (see dorsal views of endocasts). We calculated much higher values of both metrics for P. strigoides than for the other two species shown here, and the endocast of P. strigoides exhibits correspondingly visually prominent LSTCs and central expansion.

Fig. 5

The LSTC prominence and expansion ratio of vertebral canal region of the LSO across 44 bird species does not clearly correspond with terrestrial locomotion, although there is a trend of increased canal prominence in birds that perch. The scatterplots in (A) and (B) form a morphospace based on canal prominence and expansion ratio. Each point represents one species, colored by whether or not they are terrestrial or perching, respectively. (C) and (D) plot the corresponding cumulative density functions for LSTC prominence, while (E) and (F) are the corresponding histograms for LSTC prominence. Kolmogorov–Smirnov tests confirm a trend of more prominent LSTCs in perching versus non-perching species (D = 0.556, P = 0.003) but not terrestrial versus non-terrestrial (D = 0.3, P = 0.286).

Fig. 6

Locomotor mode and LSO osteological morphology are both convergent across Aves. Our interspecific sample is shown on a genus-level phylogeny from Cooney et al. (2017) based on Prum et al. (2015) and Jetz et al. (2012). Darker colors in the boxes to the right of the phylogeny indicate that the species perches or is terrestrial. Lighter colors indicate that the species does not perch or is not terrestrial. The computed LSTC prominence is shown on the bar chart.

Fig. 7

Multiple samples of five bird species occupy different areas of LSO osteological morphospace, indicating that interspecific variation in LSO osteological morphology is greater than intraspecific variation.