Distribution of Olfactory and Nonolfactory Surface Area in the Nasal Fossa of Microcebus murinus: Implications for Microcomputed Tomography and Airflow Studies

Authors


Abstract

The nasal fossa of most mammals exemplifies extreme skeletal complexity. Thin scrolls of bone (turbinals) that both elaborate surface area (SA) and subdivide nasal space are used as morphological proxies for olfactory and respiratory physiology. The present study offers additional details on the nasal fossa of the adult mouse lemur (Microcebus murinus), previously described by Smith and Rossie (Smith and Rossie [2008]; Anatomical Record 291:895–915). Additional, intervening histological sections of the specimen were used to map and quantify the distribution of olfactory and nonolfactory mucosa on the smaller turbinal of the frontal recess (FR; frontoturbinal) and those that occur between ethmoturbinals (ETs; interturbinals). A second adult Microcebus specimen, available as a dried skull, was scanned using microcomputed tomography (microCT) and reconstructed to infer the position of these turbinals within the nasal airway. Overall, turbinal bones comprise more than half of internal nasal SA. All ETs combined comprise about 30% of total nasal fossa SA, and contribute nearly half of all olfactory SA. Of these, the nasoturbinal (NT) is most completely covered with olfactory mucosa, whereas ET I is least covered with olfactory mucosa. The FR contributes significantly to total olfactory SA (ca. 20%). This recess and the single frontoturbinal within it lie in a more lateral pathway of airflow compared with interturbinals, which lie in more central zone just anterior to the olfactory recess of Microcebus. Variations in the turbinals and recesses that complicate central and paranasal in primates should be investigated further in light of zone-specific distributions of olfactory receptors (ORs) that differ between these regions in rodents. Anat Rec, 2011 © 2011 Wiley-Liss, Inc.

In a recent article (Smith and Rossie, 2008), the development of the nasal fossa in cheirogaleid primates (mouse and dwarf lemurs) was described, including estimates of internal nasal mucosal surface areas (SAs) in a perinatal and adult mouse lemur (Microcebus murinus). Generally, this study revealed that snout length is a misleading proxy for olfactory capabilities in primates, because the anterior portion of the snout disproportionately bears nonolfactory mucosa, as in many other mammals (Negus, 1958). The present study is intended to provide a more detailed analysis of olfactory mucosa distribution in the nasal fossa of Microcebus, and to place these anatomical findings into a comparative and functional context.

To clarify physiological roles of the turbinals, this study provides updated estimates of SA of all turbinals comprising elements of the adult ethmoid bone, including “accessory” turbinals that were previously not measured as separate elements (e.g., interturbinals). This enables a more detailed assessment of the total SA produced by turbinals, and the proportion to which they augment the nasal airway or its recesses. This information provides crucial background for understanding the intricate internal nasal skeletal anatomy that has recently become far more accessible for study with improved technology such as high resolution computed tomography (CT) (e.g., Rowe et al., 2005; Rossie, 2006; Smith et al., 2010; Van Valkenburgh et al., 2011).

In addition, the combined CT and histological data allow us to construct a three-dimensional (3D) model of the internal nasal skeleton in this species. Such models illustrate the functional relationship between airflow patterns and mucosal distribution within the nose.

MATERIALS AND METHODS

The sample included two adult specimens of M. murinus. One 4-year-old male mouse lemur was used to obtain measurements of SA for internal nasal structures. Smith and Rossie (2008) previously reported total internal nasal SA for this specimen, and SA for selected turbinals. In this study, additional sections were mounted to obtain more detailed estimates of SA for internal nasal structures and spaces. The nasoturbinal (NT) and interturbinals, which were not previously measured as separate structures (but were included in total SA by Smith and Rossie, 2008), were then studied. Additional sections were also examined near end points of all other internal nasal structures and spaces. This specimen was previously paraffin embedded and serially sectioned at 10 μm. Sections were stained with various procedures including Gomori trichrome, hematoxylin–eosin, and Alcian blue–periodic acid–Schiff. A second adult specimen was available for the study using previously acquired CT scan data. This specimen is a dry skull (DUPC #098, no locality info) and was scanned at the University of Texas High-Resolution X-Ray CT Laboratory, University of Texas at Austin. Data were stored as 1024 × 1024 pixels, 16-bit TIFF images. Z-spacing, or slice thickness, was set at 0.0441 mm. The skull was scanned in the sagittal plane, oriented approximately perpendicular to Frankfurt horizontal.

Three-Dimensional Reconstruction of Internal Aspects of the Nasal Fossa and Airway

For the reconstruction of internal soft tissue features of the nasal fossa, the serial sections of each nasal chamber were digitally photographed using a Leica stereomicroscope at low magnification. The digital images were then aligned for the reconstruction as described in detail by Smith et al. (2004). The reconstructions of the nasal wall were made using Scion Image software (release 4.02, NIH). To illustrate the lateral nasal wall, the septum was manually erased throughout the sequence of images using Scion Image. The reconstructions were accomplished using three different starting sections (see Fig. 1) to view the nasal fossa contour at different coronal cut-planes. Images generated from these reconstructions were saved and selected images were traced and rendered by hand to clearly emphasize contours while excluding artifactual imperfections of the sectioned tissues (e.g., folds or tears) that distort parts of the image.

Figure 1.

3D reconstructions of mucosal surfaces of the nasal fossa in an adult Microcebus, rendered after computer-generated reconstructed “stacks” of histological sections. Three coronal section cut-planes were selected (a) to demonstrate the compartments of the lateral recess and the smaller turbinals. b: The anterolateral recess (ALR), the anterior part of the lateral recess that parallels the main (central) nasal chamber. It is separated from the latter by the semicircular crest (SCr) and its mucosal extensions. The SCr is a direct ventral continuation of the nasoturbinal (NT). c: The beginning of the posterolateral recess. This posterior part of the lateral recess is divided into a dorsal frontal recess (FR) and ventral maxillary recess (MR) by the horizontally positioned frontomaxillary septum (FS). The FR communicates with the dorsal part of the main nasal fossa between the NT and ET I. d: The posterior end of the posterolateral recess, where small culs-de-sac of the FR and MR border the dorsal and ventral limits of the central nasal chamber. Here, the central nasal chamber is at its most complex, with two interturbinals (white arrows) intervening between ETs II and III. ET IV is the only major structure housed in the olfactory recess (OR). FT, frontoturbinal; MarT, marginoturbinal; MT, maxilloturbinal; N, nasal bone; ND, nasopharyngeal duct; P, palate; Pmx, premaxillary bone; SP, soft palate; and TL, transverse lamina.

3D reconstructions from CT slices were accomplished with Mimics v. 13.1 (Materialise, Leuven, Belgium). A total of 345 slices were imported with a slice spacing of 0.0441 mm and a pixel size of 0.034 mm. Airway- and skeletal-volumes were created by applying grayscale thresholds and manually editing individual slices. From these slices, high-fidelity 3D models of both the airway and skeletal components were created using Direct3D rendering and “optimal” quality.

Quantification of Olfactory and Nonolfactory Surface Area and Other Measurements

Surface area measurements.

Methodological specifics concerning identification of olfactory and nonolfactory epithelium were described previously (Smith and Rossie, 2008). Briefly, features such as Bowman's glands, multiple rows of round nuclei of olfactory receptor neurons, and olfactory nerves were used to identify olfactory mucosa. At the margins, the transition to ciliated respiratory epithelium was identified under high magnification (at 400–630× using a Leica DMLB photomicroscope, Leica Microsystems: Wetzlar, Germany).

Low magnification micrographs were used to annotate and measure olfactory and nonolfactory mucosae. Previously, individual files for every 20th section were marked according to fiducial landmarks (specific structures or contours) that corresponded to the limits of the olfactory mucosa (Smith and Rossie, 2008). In the present study, additional intervening sections were measured to improve precision for smaller turbinals (e.g., interturbinals). Excluding folded or damaged regions, every 10th section was measured. In cases where stained sections near the end points of turbinals were unusable (e.g., because of folding of tearing), unstained sections were retrieved for mounting and staining.

To obtain an estimate of olfactory SA, the perimeter of olfactory epithelium was measured in each section after calibrating to a digital image of a stage micrometer photographed at the same magnification. Every perimeter measured was recorded in millimeters, and multiplied by the distance in millimeters to the next section, yielding a segment of olfactory SA between sections. All segmental measurements were then summed and recorded as total olfactory SA. The same process was repeated on the entire perimeter of the nasal fossa to obtain total SA. Olfactory SA was subtracted from total SA to obtain nonolfactory SA. In each section, this process was completed for the entire nasal fossa, and then separately for individual structures and spaces.

In anteroposterior distance (i.e., sectional series) where focal damage prevented accurate measurement of a structure or space, the section(s) was/were excluded. In such cases, the perimeter data from the sections anterior and posterior to the damaged section were used to calculate segmental SA in the intervening space. However, even sections with some tissue distortion or damage were useful for measuring SA of individual turbinals and spaces. In summary, enough sections were analyzed to measure the end points of turbinals and spaces (e.g., anterior “tips” or “roots” of turbinals) at intervals of at least every tenth section.

Because a maximum of every 10th section was used for measurements, our SA calculations are an underestimate to an unknown extent. The end point of each structure/space was considered the last section in which it was observed by microscopy. It is conceivable that each structure and space continued for a total of nine sections before ending in an adjacent mounted section. This distance (90 μm) when multiplied by the perimeter of a structure/space in its last section, represents the maximum potential error in the SA calculations. We assessed this maximum potential error for the entire nasal fossa and for a smaller structure, the frontoturbinal.

RESULTS

Distribution of Olfactory SA in the Nasal Fossa and Its Major Components and Spaces

The addition of measurements of new sections to the previous measurements (Smith and Rossie, 2008) resulted in the addition of more SA to tips of turbinals, limits of recesses, and to the limits of the vestibule and olfactory recess. This yielded an increased overall estimate of nasal fossa SA (383.6 mm2 in this study, vs. 372.6 mm2). However, on the basis of the overall olfactory SA (117.3 mm2), the percentage of olfactory SA in the entire nasal fossa remained the same as previously reported, at 30.6 (Table 1). Updated measurements for the lateral recess remain nearly the same, at 44% olfactory SA (Table 1). Measurements of the olfactory recess remain similar to previous results, indicating it is lined mostly (69.3%) with olfactory mucosa. However, this only comprises 16% of the total olfactory mucosa in the nasal fossa (see Table 1). The nasal fossa SA is greatly elaborated by turbinals. All individually measured turbinals, when combined, comprise more than 52% of the total nasal fossa SA (this estimate does not include the marginoturbinal or semicircular crest, SCr). Detailed SA measurements for individual turbinals and smaller recesses are presented below.

Table 1. SAs of the total right nasal fossa and recesses in adult M. murinus
Structure/space for SA measurementM. murinus SA (mm2)
  1. OE, olfactory neuroepithelium.

Total nasal fossa (including recesses)383.59
 Total OE in nasal fossa117.28
 Non-OE266.31
 %OE30.57
Olfactory recess26.84
 OE in recess18.59
 %OE69.3
Nasal fossa, excluding olfactory/cupular recess (includes nasopharyngeal duct)356.75
 OE98.69
 %OE27.7
Lateral recess, total area68.03
 OE in lateral recess29.69
 %OE43.6
ALR total area33.42
 OE in ALR6.18
 %OE18.5
FR, total area25.05
 OE in FR23.5
 %OE93.8
MR, total area9.57

Three-Dimensional Reconstructions of the Nasal Fossa and Airways and SA of Specific Structures and Spaces

The reconstructions of the nasal mucosa and airways at cut-planes (Figs. 1, 2) reveal details of capacious paranasal spaces in the middle third of the anteroposterior axis of the nasal cavity. Some of these details have been previously described in strepsirrhines (Kollmann and Papin, 1925; Smith and Rossie, 2008). 3D models of the airway superimposed over skeletal components (Figs. 2, 3) emphasize the spatial relationship of the central1 nasal airway (light blue) and paranasal (dark blue) spaces. This association may also be examined in other planes in a movie file that shows this model in 360-degree rotation (available at the end of the online version of this article). Below, specific nasal structures are discussed with reference to their position in the spaces as well as their contribution to total internal nasal SA.

Figure 2.

3D reconstructions of paranasal (dark blue) and central nasal (light blue) chambers in an adult Microcebus, with the skull ghosted. Three coronal section cut-planes were selected (a) to approximately match the cut-planes shown in Fig. 1b–d. b: The ALR, c: the beginning of the posterolateral recess (two parts: FR, frontal recess; MR, maxillary recess), and d: the posterior end of the posterolateral recess. The orbits are adjacent to the paranasal spaces and the olfactory recess (OR).

Figure 3.

Cross-sectional plane (a) and 3D reconstruction (b) at a cut-plane near the anterior limit of the frontal (FR) and maxillary (MR) recesses. ET I divides the nasal fossa diagonally along a boundary (dashed line) between the central nasal chamber and the dorsolateral FR. FT, frontoturbinal.

The lateral recess is divided into two anteroposterior portions, as defined previously (Smith and Rossie, 2008). The anterolateral recess (ALR; Figs. 1b, 2b) parallels the entire dorsoventral extent of the central chamber. It is mostly separated from the central chamber by the SCr and the root of the maxilloturbinal. The separation is completed by a mucosal septum that merges the SCr with the maxilloturbinal (Fig. 1b). Most of the ALR is lined with respiratory mucosa, including especially thick compound glands ventrally. The recess only communicates with the nasal fossa posteriorly, just anterior to the connection of ethmoturbinal (ET) I to the frontomaxillary septum (FS; Fig. 1c). The posterior end of the recess contains a small amount of olfactory mucosa dorsally, comprising ∼18.5% of the ALR SA (Table 1).

The posterolateral recess begins at the first coronal plane where the FS dorsoventrally divides the lateral recess into (approximately) two halves (shown in Figs. 1c, 2c). The dorsal half, the frontal recess (FR), communicates anteriorly with the dorsal part of the nasal fossa via a narrow passageway between the first ET and the NT (Fig. 1c). These turbinals fuse posteriorly, enclosing the FR as a cul-de-sac (Figs. 1d, 2d). Throughout its length, the SA of the FR is internally complicated by a single, large frontoturbinal, and by an internal free-projection from the NT (Fig. 1c,d). The entire FR is lined with ∼94% olfactory mucosa (Table 1). The FR significantly contributes to the entire extent of olfactory mucosa throughout the nasal fossa (20% of total olfactory SA). Much of the olfactory SA is found along the roof and medial wall of the recess, the latter being lateral surfaces of the NT and ET I. The frontoturbinal is mostly lined with olfactory mucosa, but is also lined with ∼30% nonolfactory epithelium (Table 2; note newly measured sections yielded a somewhat increased estimated % olfactory SA for this turbinal compared with the previous report by Smith and Rossie, 2008).

Table 2. SAs of specific turbinals in the right nasal fossa in M. murinus
Structure for SA measurementM. murinus SA (mm2)
  1. ET, ethmoturbinal; FT, frontoturbinal; NT, nasoturbinal; and OE, olfactory neuroepithelium.

ETI SA40.47
 ETI OE13.89
 Non-OE26.58
 %OE34.3
ETII SA13.23
 ETII OE7.16
 Non-OE6.07
 %OE54.1
ET III SA10.57
 ET III OE6.74
 Non-OE3.83
 %OE63.8
ET IV SA20.03
 ET IV OE9.03
 Non-OE11
 %OE45.1
Maxilloturbinal SA78.45
Interturbinals 1 and 2 SA10.81
 OE4.61
 Non-OE6.2
 %OE42.6
FT SA11.72
 FT OE8.16
 Non-OE3.56
 %OE69.6
NT SA7.41
 NFT OE7.09
 Non-OE0.32
 %OE95.7

The ventral “half” of the posterolateral recess, the maxillary recess (MR), is internally much less complex. A small lateral projection from the SCr (not shown) is the only osseous elaboration of the recess wall. Correspondingly, the total SA of the MR is less than half that of the FR, and none of it is lined with olfactory mucosa (Table 1).

At the posterior limit of the posterolateral recess, the two parts are displaced dorsal and ventral to the main, central nasal cavity (Figs. 1d, 2d). Here, just anterior to the olfactory recess, the ET complex is complicated by the addition of accessory scrolls between ETs II and III (Fig. 1d). One of these scrolls is partially attached to the root of ET III (not shown) before diverging to anchor into the lateral wall. Where attached to ET III, the structure is consistent with Maier's (1993) definition of an “epiturbinal,” an accessory scroll of the ET itself. Posterior to this attachment, this scroll is more extensively anchored to the lateral wall of the nasal fossa. Thus, we consider this entire structure to be the second of two “interturbinals,” small scrolls from the lateral wall of the pars posterior that do not reach the midline (Smith and Rossie, 2008). The two interturbinals, in summary, are lined with ∼43% olfactory SA.

The SA measurements for other turbinals are similar to the previous measurements; only minor differences between the updated and previous measurements of SA of the maxilloturbinals and ETs I and II can be found if Table 2 is compared with the results in Smith and Rossie (2008, Table 3 therein). New measurements made in the present study yielded an increased estimate for the percentage of olfactory SA on ET III compared with previous findings (63.8% vs. 57.3%). Measurements of olfactory and nonolfactory SA on ET IV are each increased compared with previous results, but the % olfactory SA remains nearly the same (Table 2). The measurements of the NT, which were not previously reported, indicate that this structure is covered by a greater extent of olfactory SA (95.7%) than any other turbinal (Table 2).

When all structures/spaces are assumed to continue anterior and posterior 90 μm and anterior to the first and last section in which they appear, the calculation of total SA in the nasal fossa would increase to 389.16 mm2. When compared with our estimate, this would represent a 1.4% error. The maximum error for smaller structures is calculated to be larger (e.g., maximum error for the frontoturbinal is ∼3.1%). We believe the actual error to be smaller for two reasons. First, it is unlikely that every space/structure continues in all sections after its start or stop point. Second, most structures decrease in cross-sectional area at their anteroposterior end points; our maximum error calculation does not account for this possibility.

DISCUSSION

Numerous reports have provided olfactory SA in mammals. In almost all reports, the data are a sum of all internal nasal surfaces and not the component parts (e.g., turbinals). The present study, in an effort to more precisely estimate SA for nasal turbinals and spaces in Microcebus, found the same cumulative 30.6% of the total nasal fossa SA is lined with olfactory neuroepithelium as in a previous report (Smith and Rossie, 2008). Few reports on olfactory SA exist for other primates (e.g., Woollard, 1925; Smith et al., 2004). No reports on the percentage of olfactory SA exist for nonhuman anthropoids, although they would appear to have a lesser percentage than Microcebus (Smith, unpublished observations). The adult Microcebus measured here has a lesser percentage of olfactory SA compared with Gurtovoi's (1966) report on rodents such as Clethrionomys (50%) and Microtus (37%) or insectivores such as Sorex (69%) or Erinaceus (47%). Other reports on rodents indicate ∼54% of the nasal fossa in a mouse (genus Peromyscus) and 41% of the nasal fossa of the hamster (Mesocricetus auratus, Clancy et al., 1994) is olfactory. Two reports show that bats (Chiroptera) may have a similarly wide range in percentage of olfactory SA (25%–42%, Gurtovoi, 1966; 29%–56%, Bhatnagar and Kallen, 1975).

Taken together, the existing comparative data suggest even the “macrosmatic” primate Microcebus may not match the percentage of olfactory SA seen in insectivores, rodents, and some bats. These data must be interpreted with caution, as SA measurements by themselves are imperfect proxies for olfactory neuron population size. For example, SA does not take cell size or density into account (Smith and Bhatnagar, 2004). However, SA measurements do provide physiological insight into the roles of individual nasal elements, such as the turbinals (Smith et al., 2007a, b). Therefore, our knowledge of the specific distribution of mucosae over these elements informs comparative analyses in a crucial manner. A broader base of histological data on these structures has become more important as high resolution study of the internal nose is now possible using CT technology (Rowe et al., 2005; Craven et al., 2007; Smith et al., 2010). The present study offers a framework for future quantitative work. These observations also allow some preliminary comparative statements on the nasal fossa elements.

Comparison to Internal Nasal Dimensions in Monodelphis domestica

Despite obvious methodological differences between the present study and Rowe et al. (2005), both are rooted in mucosal characteristics of nasal elements, allowing some comparative statements. There is a close similarity in the organization of the nasal fossa in two respects. By each estimate, the maxilloturbinal comprises approximately one-fifth of total nasal fossa SA. Also, in terms of the division of labor vis-à-vis olfactory and respiratory function, both studies attribute dual function to the first ET (see also Smith et al., 2007a). However, Rowe et al. assign all ETs (using the term “endoturbinals”) caudal to ET I to solely an olfactory function, whereas this study indicates only closer to one-half of the SA of the posterior ETs is covered with olfactory neuroepithelium. Whether this reflects a significant difference between taxa deserves further study using histology. The mucosal distribution in Monodelphis was estimated based on a comparison of adult CT slices to histology from subadults. If Monodelphis has a disproportionate age-related increase in nonolfactory mucosa, as observed for primates (Smith et al., 2007a), the mucosal estimates for the ETs may somewhat overestimate the amount of olfactory mucosa.

Regardless of such differences, both studies agree that the ET complex (here, meaning all ETs in the medial row, interturbinals, frontoturbinals, and the osseous part of the NT) bears substantial quantities of olfactory neuroepithelium. In this light, some differences between Microcebus and Monodelphis are particularly striking. For example, the SA for ETs in Monodelphis (“endoturbinals” in Rowe et al., 2005) comprises 37% of the total nasal fossa SA, whereas this percentage is 25% in Microcebus (this percentage includes both interturbinals and ETs in Microcebus, because Rowe et al. did not distinguish the former). Rowe et al. also provide SA for two “ectoturbinals.” As these are located superior to the root of ET I, or the FS (see Fig. 6 in Rowe et al., 2005), these correspond to “frontoturbinals” as defined here (Smith and Rossie, 2008). Respectively, the frontoturbinals comprise 10% and 3% of total nasal fossa SA for Monodelphis and Microcebus. The NT comprises 3.5% of total nasal fossa SA in Monodelphis. A comparison of SA for the NT is less straightforward, because Rowe et al. include a portion of bone that projects anterior to the FR (see Fig. 5 in Rowe et al., 2005). This portion corresponds to the SCr in Microcebus (see Fig. 1b). When the SCr is combined with the NT of Microcebus, the cumulative SA (11.36 mm2) comprises ∼5% of total nasal fossa SA. In summary, the ET complex of Monodelphis comprises 50% of total nasal fossa SA (Rowe et al., 2005; Table 2), whereas this percentage is considerably less in Microcebus. Even if the nasal vestibule is excluded from measurements of Microcebus (it was not included by Rowe et al., as it extends outside the confines of the nasal skeleton) and the SCr is included, the ET complex only comprises 33% of total nasal fossa area.

Aside from the SA for the NT, which is relatively similar, these comparisons indicate organizational differences in the extent to which turbinals augment internal nasal surfaces. On the basis of the available data, both the FR and all posterior spaces are packed with a greater percentage of turbinal SA in Monodelphis than in Microcebus.

Significance of the Distribution of Olfactory SA to Airflow

Our knowledge of internal nasal airflow is based mainly on humans, rodents, and carnivores (Settles, 2005; Zhao and Dalton, 2007; Craven et al., 2010). Craven et al. (2010) relate the development of an olfactory recess to unique airflow patterns in “macrosmatic” mammals such as carnivores and ungulates. At resting rates of respiration in canines, most (nearly 90%) inspired air flows ventral to the conduit by which odorants reach the ethmoidal region, the dorsal meatus. Sniffing appears to deliver odorants to the olfactory recess and other chambers via the dorsal meatus, thereby bypassing the convoluted respiratory passages below (Craven et al., 2010). Microcebus presents an interesting primate model with which to examine the relationship between airflow patterns and anatomy because the olfactory recess houses only one of the posterior ETs (Smith and Rossie, 2008). Correspondingly, the olfactory recess may be small compared with many other mammals, such as the dog, where most of the ET complex resides within the enclosed recess. Indeed, olfactory recess SA comprises only 7% of the total nasal fossa SA in Microcebus. Future comparisons to other mammals may reveal whether Microcebus possesses a proportionally small olfactory recess, and whether this is a primitive condition for primates.

Another aspect of nasal architecture emphasized by the present study requires a broad comparative study for a better physiological understanding: the significance of smaller turbinals. As described by Smith and Rossie (2008), turbinals develop in two regions of the developing nasal capsule, the pars intermedia and the pars posterior. In postnatal morphology, these regions correspond to skeletal support for the paranasal region and the central portion of the ethmoid labyrinth respectively, although the ethmoid bone itself forms boundaries for both of these. The central portion of the ethmoid labyrinth is directly continuous, posteriorly, with the olfactory recess. As interturbinals occur between ETs II and III in Microcebus, they augment a spatially more centralized airspace than frontoturbinals, which occur only in the dorsolaterally restricted FR. Thus, the distinction made by some authors (Maier, 1993; Smith and Rossie, 2008) among these smaller “ectoturbinals” may be more than just one of developmental anatomy.

The positional differences between frontoturbinals and interturbinals take an added significance when considering olfactory axon pathways in rodents. In the hamster, the anterior, free extension of ET I provides a somewhat vertical plate that divides airflow into lateral and more medial (centralized, toward the olfactory recess) directions (Clancy et al., 1994; Schoenfeld et al., 1994). Microcebus shares this anatomical arrangement (Fig. 3). The lateral region in the hamster (termed “lateral olfactory recess” by Clancy et al.) is nearly equivalent to the paranasal FR, as termed here. Even in well-studied mammals, such as the dog, airflow in paranasal spaces is poorly understood compared with the main airway (Craven et al., 2007). But evidence on the nature of receptivity of the olfactory neuroepithelium indicates these two regions may not detect an identical range of odorant types. In rodents, the OR neurons located along more central ETs project axons to spatially distinct portions of the main olfactory bulb compared with those along frontoturbinals (a.k.a., “ectoturbinals”) (Miyamichi et al., 2005). In addition, different populations of OR neurons (which express specific OR genes) are distributed zonally along positions corresponding to more central (along dorsal meatus) or more peripheral paths of inspiratory airflow (Schoenfeld and Cleland, 2005, 2006). Although most zones of OR neuronal populations appear to overlap to some degree (Miyamichi et al., 2005; Schoenfeld and Cleland, 2005), the internal complex of nasal fossa of rodents is thus highly organized with regional specializations in odorant detection. Airflow modeling studies have shown that odorants with different chemical properties (e.g. molecular weight and solubility) may deposit in different regions of the nasal passages, and that these regions of deposition appear to correspond to the zones of OR expression (Yang et al., 2007). Schoenfeld and Cleland (2005) suggest that this design, in which the distribution of receptors across internal nasal space aids in odorant discrimination, may be a characteristic of long-snouted vertebrates. But what if these spatial characteristics of OR neuronal distribution are but one nuance in a broader vertebrate trend in neuroepithelial organization? For example, is it possible that particular zones of OR neurons are expressed on specific turbinals, and that the appearance of these turbinals varies across mammals? If so, then morphological variations in turbinal structure and number may relate to species differences in odorant receptivity. This would have important implications in the study of primate nasal anatomy, where a complex trend in nasal fossa reduction may typify some or all of the order (Smith et al., 2007b; Smith and Rossie, 2008). Unfortunately, knowledge of intranasal airflow in primates is heavily biased toward monkeys and humans, in which a relatively turbulent flow pattern predominates during sniffing behaviors, whereas airflow in rodent nasal fossae occurs in a laminar pattern (even during sniffing behaviors) which allows diffusion of odorants across peripheral spaces (Schoenfeld and Cleland, 2005). The majority of “long-snouted” primate species, such as strepsirrhines, remain uninvestigated in regard to intranasal airflow patterns.

CONCLUSION

Highly detailed information on the distribution of nasal mucosa types, as well as intranasal airflow patterns, is presently available for relatively few mammalian species. The present study begins to address the need for further studies by quantifying olfactory and nonolfactory components of the nasal wall in a nocturnal strepsirrhine, a representative of a taxon that may be ecologically and behaviorally similar to early primates (Cartmill, 1974). The overall percentage of olfactory mucosa in M. murinus is less than that in many other groups of mammals, despite its numerous ETs. However, olfactory mucosa is distributed across a broad array of structures and passageways within the nasal fossa and paranasal spaces. This may indicate that in primates such as Microcebus, the distribution of olfactory mucosa is reorganized to facilitate deposition of odorants during sniffing behaviors. A thorough understanding of nasal cavity architecture and function will require knowledge of airflow patterns as well as the spatial distribution of these sensory structures.

Primates are characterized by a reduction of the nasal fossa relative to other mammals, although the extent varies greatly across the order (Smith et al., 2007b). Because of varying degrees of orbital convergence toward the midline, the FR is nearly or completely obliterated in haplorhines (Cartmill, 1970; Smith et al., 2007b). It has been observed that the FR may also be somewhat “compressed” by enlarged orbits in at least some strepsirrhines compared with most other mammals (Smith and Rossie, 2008). Such changes were surely accompanied with alterations to intranasal airflow patterns, but these have yet to be investigated. A more specific question is how the spatial distribution of odorant receptors across the primate nasal fossa is affected by structural reduction. A broader understanding of associations between turbinal types and OR types among mammals appears to be required first. If a predictable relationship exists across mammals, then the significance of specific structural reductions in primates will become far clearer.

Acknowledgements

We thank ER Dumont (University of Massachusetts Amherst) for access to computer resources to generate the 3D volumes based on CT scans. This is Duke Lemur Center Publication Number 1198.

  1. 1

    Terminology for regions and structures of the nasal fossa vary enormously in the existing literature (see Moore, 1981; Smith and Rossie, 2006). Frequently, terminology is rooted in a physiological (e.g., Clancy et al., 1994; Craven et al., 2007) or developmental (e.g., Maier, 1993; Rowe et al., 2005; Smith and Rossie, 2008) basis. Here, we use the term central nasal airway primarily in a descriptive manner. It is not equivalent spatially with the central nasal chamber per Clancy et al. (1994). Rather, it is meant to denote medial airway regions as opposed to the more lateral paranasal spaces.

Ancillary