SEARCH

SEARCH BY CITATION

Keywords:

  • infraorbital foramen;
  • infraorbital nerve;
  • infraorbital artery;
  • vibrissae;
  • paleoecology

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Osteological cranial features, such as foramina, assist in phylogenetic and ecological interpretations of fossil mammals. However, the validity of using foramina in these interpretations when their contents are not well documented is questionable. For decades, the infraorbital foramen (IOF) has been used to interpret aspects of the fossil record, yet there are conflicting accounts about what passes through the foramen and little known about how neural and vascular structures contribute to its contents. This study tracks and documents the neural and/or vascular anatomy of the IOF and examines the correlation of infraorbital nerve (ION) and IOF cross-sectional area. To address this question, 161 mammalian cadavers, including 80 primates, were injected with latex dye to track the vascular anatomy associated with the IOF. All ION fibers were then removed from the infraorbital canal, and ION cross-sectional area was calculated from histological slides. Latex injections and histological slides revealed that only the ION and a small infraorbital artery pass through the IOF. Variation in ION size explains 85% of variation in IOF area, and the artery represents a negligible portion of the foramen. The strong positive correlation between the ION and IOF size suggests that, in the absence of nerve tissue, the IOF can serve as a proxy for ION area. IOF area maybe used to evaluate differences in maxillary mechanoreception in both extinct and extant taxa. Anat Rec, 291:1221–1226, 2008. © 2008 Wiley-Liss, Inc.

Infraorbital foramen (IOF) is found below the inferior margin of the orbit, and for decades, its relative cross-sectional area (area) has been used in ecological and phylogenetic interpretations of the fossil record (Kay and Cartmill, 1977; Gingerich, 1981; Gingerich and Martin, 1981; Zheng, 1984; Simons 1987, 1997; Czaplewski, 1991; Kay et al., 1992, 2004a,b; Rasmussen and Simons, 1992; Fleagle, 1999; Brunet et al., 2002; Shigerhara et al., 2002; Ni et al., 2004 ; Beard and Wang, 2004; MacPhee and Horovitz, 2004, Rossie et al., 2006; Goin et al., 2007). These interpretations are based on the underlying assumption that IOF area reflects the sensory acuity of the maxillary region, because the afferent infraorbital nerve (ION), which passes through the IOF, innervates mechanoreceptors of the maxilla (Patrizi and Munger, 1966; Gasser and Wise, 1971). For example, IOF area has been linked with the number and “development” of vibrissae (Kay and Cartmill, 1977; p.42) and the presence or absence of a rhinarium (Gingerich, 1981; Rossie et al., 2006). Others have proposed that IOF area reflects differences in global temperature climes, where populations in colder climates are hypothesized to have larger infraorbital arteries to warm the face, and therefore larger infraoribital foramina, than equatorial populations (Churcher, 1959). These hypotheses are based on the assumption that IOF area is closely correlated with ION or infraorbital artery area. To date, it is unknown to what degree these structures make up the contents of the infraorbital canal.

In cases where the relationship between the contents of a canal or foramen and its area are unknown, there has been controversy about the validity of using these features to interpret the fossil record (Kay et al., 1998; Degusta et al., 1999; Jungers et al., 2003). For example, Kay et al. (1998) proposed that the size of the hypoglossal canal might help pinpoint the first appearance of humanlike speech in the fossil record. The hypoglossal canal transmits CNXII, which is responsible for motor control of the tongue. Compared with other mammals, humans have relatively larger hypoglossal canals, and it was assumed that its large size in humans reflects hypoglossal nerve area, because speech requires finer motor control of the tongue (Kay et al., 1998). However, detailed dissections documented that a venous plexus is the largest constituent of the hypoglossal canal (Jungers et al., 2003). Conversely, optic foramen area, in association with relative eye size, is an established osteological proxy for retinal summation, because the optic nerve and foramen area are highly and significantly correlated (r2 = 0.88, P < 0.0001; Kirk and Kay, 2004).

Presently, there are conflicting observations in the mammalian and human anatomical literature about what vessels run through the IOF, and little is known of how the soft anatomy affects its size. Some texts report that the ION, an infraorbital artery, and a vein are transmitted through the IOF (Hoskins and Lacroix, 1949; Stern, 1987; Smith, 1999). Other anatomical works suggest that only the ION and infraorbital artery exit the foramen (Gray, 1995; Done et al., 1996; Moore and Dailey, 1999).

This study has two objectives: to gain a better understanding of the anatomy of the infraorbital canal and to determine how its contents vary with the area of the IOF. The vascular anatomy of the IOF can be examined in mammalian cadavers using latex injections. Accurate measures of vessel area are not possible, because vessels collapse after death and the latex injections artificially inflate vessels size. However, ION area can be calculated from cadavers and used to determine the correlation between IOF and ION area. Until we establish which anatomical structures, if any, correlate with IOF area, we cannot use the IOF in ecological interpretations of the fossil record.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Sample

One hundred sixty-one fresh frozen mammalian cadavers, from seven mammalian orders, were used for this study (Table 1). Cadavers were prepared for dissection by injecting dyed latex into the left or right common carotid artery (RED) and internal jugular vein (BLUE). Latex injections allowed for clear vascular tracking. Each specimen's maxillary region was then dissected to qualitatively identify vascular anatomy.

Table 1. Sample and infraorbital nerve and foramen measurements
OrderSpeciesNION cross-sectional area (mm2)IOF cross-sectional area (mm2)
MeanMinMaxSDMeanMinMaxSD
ArtiodactylaCapra hircus117.192.9311.442.6512.018.2015.492.44
 Odocoileusvirginianus411.545.9618.285.1212.7610.6315.562.38
 Sus scrofa525.4322.2029.583.0731.3526.2237.005.01
CarnivoraCanis familiaris87.682.7414.93.8814.248.2918.563.85
 Canis lupus27.566.039.092.1618.3515.1621.534.50
 Enhydra lutris214.3414.0914.580.3532.8429.4536.224.79
 Felis silvestris134.631.0825.536.369.293.7420.774.85
 Lynx rufus13.693.693.69 9.109.109.10 
 Procyon lotor54.070.746.532.4213.991.7126.2210.50
 Ursus americanus213.8311.4316.223.3920.9119.5222.291.96
 Vulpes vulpes33.591.655.631.996.735.238.021.41
 Zalophus californianus15.805.805.80 13.1513.1513.15 
DidelphimorphiaDeidelphus virginianus35.122.466.742.325.983.068.452.72
InsectivoraPeromyscus californicus21.190.342.041.201.751.062.430.97
 Scapanus townsedii30.200.200.200.001.391.341.440.07
PerissodactylaEqus caballus731.9123.0547.9511.1934.0419.5853.4113.17
PrimateCallithrix jacchus20.390.330.450.080.530.470.580.08
 Callifthrix gygmaea10.850.850.85 0.860.860.86 
 Cebus apella41.120.971.220.121.221.101.300.10
 Cercopithecus albogularis31.750.792.701.351.321.171.470.21
 Cercopithecus mitis30.460.270.640.260.880.511.250.52
 Chlorocebus aethiops101.250.282.740.861.580.712.740.68
 Erythrocebus pata41.911.252.490.612.121.203.541.00
 Gorilla gorilla211.0010.1311.871.2311.9510.4713.432.09
 Homo sapiens33.192.334.051.225.034.415.640.87
 Loris tardigradus40.300.180.480.130.380.260.490.09
 Macaca fuscata91.550.483.291.072.020.823.440.96
 Macaca mulatta41.790.892.730.761.841.043.431.08
 Macaca nemestrina22.161.672.650.691.581.181.970.56
 Mirz coquerli30.650.470.830.252.002.002.000.00
 Nycticebus coucang20.390.260.510.180.600.580.620.03
 Otolemur crassicaudatus21.020.911.130.161.681.51.860.25455844
 Pan troglodytes63.851.007.812.794.961.948.522.63
 Papio hamadryas76.291.3318.666.356.791.3222.417.25
 Perodicticus potto20.870.401.330.661.180.621.740.79
 Varecia variegata51.090.621.410.391.751.002.960.85
 Chlorocebus sabaeus20.970.761.180.301.231.061.390.23
RodentiaRattus rattus410.631.2321.1510.2016.455.4028.9612.70
 Sciurus niger51.231.051.410.251.531.271.790.37

Once the gross anatomy of the IOF region was described, the entire contents of the infraorbital canal were removed to create histological slides. Histological slides were used to calculate ION area and to confirm the presence or absence of vessels (Fig. 1). During the removal of the contents of the infraorbital canal, the exact location where the ION exited the infraorbital canal through the IOF was marked. The nerve bundles were then placed in 10% formalin. An automated tissue processor dehydrated and infiltrated the tissues, which were then embedded in paraffin. The paraffin blocks were sectioned transversely at right angles to the long axis of the nerves and vessels using a Shandon Finesse 325 Microtome® at 5-μm intervals along the nerve. Ten serial sections were mounted on slides, and the tissue was stained using hematoxylin and eosin. Each serial section was photographed under a microscope, and the images were used to calculate nerve area (mm2) in Scion Image® software. Because slight folds and wrinkles in paraffin can alter the calculated measurement of ION area, the mean of 10 serial sections was calculated to resolve slight differences in ION area measurements due to artifact or interobserver error. Although the exact location of where the ION exited the IOF was labeled, the mean of 10 serial sections ensures accurate coverage to offset any error that may have occurred during sectioning. The above-described methods have been tested and applied to address similar questions (DeGusta et al., 1999).

thumbnail image

Figure 1. A photograph of a histological coronal cross section of the tissue contents of the infraorbital canal of a domesticated dog. A labels an artery and N labels nerve bundles. The remaining unlabeled tissues surrounding the nerve bundles and artery are either fat or blood. Histologically, arteries are easily distinguished from veins because arteries have thicker walls.

Download figure to PowerPoint

During tissue processing, the ION shrinks, but dehydration of the nerve should not affect the investigation into how the ION and IOF correlate for two reasons. First, preparation and processing of each specimen is identical, and therefore dehydration is the same across specimens. Second, the processed nerves are compared with the foramina from which they came. The uniformity in processing and the within-specimen comparisons ensure an accurate assessment of the correlation between ION and IOF area.

The IOF is an irregularly shaped foramen and to measure this feature accurately, molds were created using flexible injectable-molding material (Coltène President Plus, Regular Body Molding Material). Molding material was injected into the canal, and when the material hardened, a mark was made on the molding material pointing to the infraorbital canal outlet. The molds were then sectioned at the outlet mark in the coronal plane and then photographed under a microscope. These images were used to quantify (mm2) IOF area in Scion Image software.

Statistical Analysis

Species means were created for IOF and ION measurements. Individual specimen values and species means were natural log transformed to normalize the data. The IOF and ION area data were compared using a Spearman's rank correlation. A least squares (LS) and reduced major axis (RMA) regression were performed to explore the scaling relationship among the variables. A LS regression of ln ION area (dependent variable) and ln IOF area (independent variable) was calculated for strepsirrhines and haplorhines, respectively, and an analysis of covariance (ANCOVA) was used to test for differences in slope and y-intercept between these groups. The ANCOVA data met model assumptions (e.g., regression slopes not significantly different). Only the results for differences in y-intercept will be presented. Significance for the ANCOVA analysis was set at P < 0.05.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Qualitative observations based on latex injections of the common carotid artery and internal jugular veins show that only an infraorbital artery and the ION pass through the IOF. None of the specimens sampled had a vein passing through the IOF. There was variation in artery size. The primates sampled had discernibly smaller arteries compared with some of the carnivores. Domesticated felids had the largest arteries. Nonetheless, compared to the ION, the artery was small compared with the nerve in all specimens sampled. The histological slides confirm these observations (Fig. 1).

A significant positive correlation (Spearman's rank: Rho = 0.91; P < 0.001) exists between individual specimens' ION and IOF areas (Fig. 2). IOF and ION species' mean data show a stronger, but nearly identical, correlation (Rho: 0.94; P < 0.001). A RMA regression analysis for both individual specimen and species mean values shows that IOF and ION area scale isometrically (Fig. 2). A LS regression indicates that there is a slight negative allometric relationship for both individual specimen and species mean data, but the confidence intervals include isometry (Fig. 2). The Townsend's mole (Scapanus townsendii) was the only outlier. This species had relatively smaller nerves than expected given the size of its foramina. The LS regression analysis using individual specimen data shows that 83% of the variation in IOF area is explained by ION area. When species means were considered in a LS regression analysis, 90% of IOF area is explained by ION area (Fig. 2). Based on the results of LS regression analysis for both individual specimen values and species means values, no determinable grade shifts were observed across the mammalian orders sampled. An ANCOVA shows that there are no significant differences between strepsirrhines and haplorhines in the relative amount of space the ION occupies within the IOF (r2 = 0.84, F(1,74) = 2.71, P = 0.10; Fig. 3).

thumbnail image

Figure 2. A scatter plot depicting the relationship between IOF and ION area. The solid and the dashed line represent least squares (LS) and reduced major axis (RMA) regression respectively. A: Figure of individual specimen values. There is a significant correlation (Spearman's Rho: 0.91; P < 0.001) between these variables. The LS (ln ION area = −0.39 + 0.94 ln IOF area; r2 = 0.83) and RMA (ln ION area = −0.52 + 1.03 ln IOF area; r2 = 0.82) regressions show an isometric relationship between the two variables. B) A graph showing a significant correlation of species mean values (Spearman's Rho: 0.94; P < 0.001) for ION and IOF area. The LS (ln ION area = −0.41 + 0.95 ln IOF area; r2 = 0.90) and RMA (ln ION area = −0.48 + 1.01 ln IOF area; r2 = 0.90) regression show that IOF and ION scale isometrically. X = Artiodactyla, • = Carnivora, Z = Didelphimorphia, Y = “Insectivora,” + = Perissodactyla, ▪ = Primates, and * = Rodentia.

Download figure to PowerPoint

thumbnail image

Figure 3. A scatter plot depicting the relationship between ln IOF and ln ION area among individual primate specimens. The solid line represents a least squares (LS) regression line for strepsirrhines and the dashed line is the regression line for haplorhines. An ANCOVA indicates that there are no significant differences between the two groups. The haplorhines are the open circles and strepsirrhines are the closed circles.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Results from this study suggest that the area of the osteological IOF can serve as a proxy for ION area. There is no vein and only a small artery present in the infraorbital canal. The size of the infraorbital artery and its possible role in heating the face during cold-induced vasodilatation has been proposed as a possible explanation for IOF area variation among mammals, including humans (Churcher, 1959). The strong correlation between the ION and IOF found in this study does not support this hypothesis. Cold-induced vasodilatation is a physiological adaptation (Lewis, 1930; Burton and Edholm, 1955), but this transient change to vessel cross-sectional area would not influence the findings of this study. Furthermore, the arteries that bring blood to the face do not go through the IOF; instead, they are located outside the foramen and are too large to pass through the IOF (Washburn, 1963).

Because the majority of foramen area can be explained by nerve area, what do differences in ION area signify? The afferent ION innervates Merkel, Ruffini, and Pacinian mechanoreceptor corpuscles located in the maxillary region (Martini and Timmons, 1997; Ebara et al., 2002). Mechanoreceptors are sensory receptors that respond to tension, pressure, and displacement. The sensory acuity as well as the cortical representation of a region is dependent on receptor density (Catania and Kaas, 1997; Dehnhardt and Kaminski, 1995; Nicolelis et al., 1997). Regions with high-receptor densities need more nerve fiber innervation (Kandel et al., 2000; Oelschlager and Oelschlager, 2002; Marino, 2007). Nerve area is a reliable estimate of total nerve axon count, as examples from total optic (Jonas et al., 1992; Cull et al., 2003) and hypoglossal (Mackinnon and Dellon, 1995) nerve area suggest. Mackinnon and Dellon (1995) documented variation in axon densities across an individual nerve, but when they calculated total axon count across nerve samples within a single taxon, they noted that axon count correlated with nerve area. Based on these studies, and because IOF and ION areas significantly and highly correlate, IOF area may be used to infer sensory acuity of the maxilla among mammals in the absence of nerve tissue.

IOF area is often used to evaluate the sensory abilities of mammals. IOF area has been used to predict both the number and “development” of vibrissae (Kay and Cartmill, 1977) and the presence or absence of a rhinarium (Gingerich, 1981; Rossie et al., 2006). For instance, animals with relatively large foramina were assumed to have many vibrissae and/or a rhinarium, whereas smaller foramina were associated with a lack of a rhinarium and fewer vibrissae. Except for the haplorhines, all mammals sampled for this study have a rhinarium. Among primates, there were no significant differences between strepsirrhines (which have a rhinarium) and haplorhines in the relative amount of space in the IOF occupied by the ION. This suggests that the IOF cannot be used to predict the presence or absence of a rhinarium. On the other hand, IOF area may predict the number and shape (thickness and/or length) of vibrissae. Mystacial vibrissae are special tactile hairs that assist in general spatial exploration and tactile object recognition tasks (Gottschaldt et al., 1973; Brecht et al., 1997; Carvell and Simons, 1990; Dehnhardt and Kaminski, 1995). The specialized mechanoreceptors previously described are located along the shaft of a vibrissa along the glassy membrane, a basement membrane surrounding the follicle. Experimental work on maxillary mechanoreceptivity fields in rats (Nicolelis et al., 1997) and pinnipeds (Dehnhardt and Kaminski, 1995) report that the object recognition correlates with receptor density and nerve thickness. Therefore, if there is an increase in vibrissa number, there is an expected increase in receptor and axon number, and as a result, ION and IOF areas are expected to increase.

Predicting the number and shape of vibrissae has implications for interpreting the ecology of mammals. There is a strong relationship between vibrissae and ecology (Vincent, 1913; Kratochvil, 1968; Gottschaldt et al., 1973; Ahl, 1987; Dehnhardt and Kaminski, 1995). For example, Kratochvil (1968) associate vibrissa number with the activity pattern, because they found that nocturnal species have more vibrissae than diurnal species. Kratochvil (1968) concludes that because nocturnal animals rely less on vision than diurnal animals, vibrissae are of greater importance to them for detecting objects that are near the face. How an animal uses its face to explore the world directly affects the number of receptors present in the face, which, in turn, affects nerve area and IOF area.

A preliminary investigation into the relationship between vibrissae and IOF area shows that vibrissa count significantly correlates with IOF area (Muchlinski, 2005, 2008). These results further support the idea that IOF area can be useful in understanding how ecology shapes maxillary sensory acuity. Depending on how animals use their faces to explore the world, the association of ecology and maxillary mechanoreception is expected to change across orders. With a correlation between ION and IOF established, attention can shift to understanding how the IOF, and mechanoreception by proxy of anatomy, is shaped by ecology. Results of this work can certainly be applied to paleoecological interpretations of the fossil record.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

I thank Liza Shapiro and Chris Kirk, for their support and help in developing this project, and Nate Dominy, Texas A&M Veterinary Hospital, The University of Texas at Austin Anthropology Laboratory, and The University of California, Santa Cruz, for cadavers. I also thank Richard Baldwin and Jamie McGee for their help in harvesting IONs from cadavers. Sample processing was completed at the M. D. Anderson Histology Laboratory in Smithville, Texas. I thank Irma Conti and her laboratory staff for teaching me histological techniques and allowing me to process my nerve samples. Editing was rendered by Becky Ham.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED