Ontogeny of the masticatory muscles in the opossum Didelphis albiventris (Marsupialia, Didelphimorphia, Didelphidae)

Abstract

Opossums (marsupials of the Didelphidae family) retain a generalized masticatory apparatus and tribosphenic molars, often used as models to understand the evolution of mastication in early therian mammals. Like all marsupials, their growth goes through a stage when pups complete their development while permanently attached to the mother's teats before weaning and starting feeding on their own. Yet, while the masticatory muscles of adults are known, as is the ontogeny of the cranium and mandible, the ontogenetic changes in the masticatory muscles remain unknown. Here we describe for the first time the changes in the masticatory muscles observed in lactating pups, and weaned juveniles, subadults, and adults in the White‐eared opossum, Didelphis albiventris, through dissection of 25 specimens and quantification of relative muscle masses, lines of actions and mechanical advantages whenever possible. We also assessed the scaling patterns of muscle masses and mechanical advantages through ontogeny. The main changes, as expected, were found between suckling and weaned specimens, although some changes still occurred from juveniles to adults. The adult adductor musculature is similar to the other Didelphis species already known, with a dominant m. temporalis that originates on the lateral wall of the skull, up to the sagittal and nuchal crests, and fills the zygomatic arch when inserting into the lateral and medial surfaces of the coronoid process, respectively through the pars superficialis and pars profunda. The m. masseter is also subdivided in superficial and deep bundles which originate posteriorly in the maxilla and zygomatic arch, and insert into the angular process and masseteric fossa in the mandible. The m. pterygoideus medialis originates from the palatine, the pterygoid bone and the alisphenoid, and it inserts on the angular process medially. Suckling pups showed muscles with more restricted attachments, reduced muscle lines of action, and less diversity in the fiber orientation. The absence of the postorbital constriction also resulted in a distinct morphology of the m. temporalis pars profunda, through two bundles, one anterior and one posterior, which insert more inferiorly into the mandible. These major changes can be related to the onset of mastication and to size‐related changes in growing weaned age classes. In general, all adductor muscles grew with positive allometry, and increased their fixation areas through, in part, the development of specific regions of the cranium and mandible. Their lines of action also increase and diversify along ontogeny. These changes can be related to the functional requirements for fixation during lactation, which shift to adduction and mastication movements after weaning.

Muscle locations and estimated lines of action of the superficial (red) and deep (orange) temporals, and superficial (dark blue) and deep masseter (light blue) in lateral view of representatives of different age stages of Didelphis albiventris (Didelphidae).

1. INTRODUCTION

The mammalian masticatory apparatus is involved in several, sometimes conflicting, functions, yet its morphology seems to be mainly influenced by feeding habits (Herring & Herring, 1974). Morphological and functional features of the masticatory apparatus are adapted to different diets (Shi et al., 2020), combining information on anatomy, mechanics, and behavior (Herring, 1993). The interaction between these characters and the performance of the masticatory system has been widely studied (Binder & Van Valkenburgh, 2000; Dumont et al., 2009; La Croix et al., 2011), particularly through biomechanical models (Brassard, Merlin, Guintard, Monchâtre‐Leroy, Barrat, Callou, et al., 2020; Fabre et al., 2017; Hartstone‐Rose et al., 2018; Leonard et al., 2022; Santana et al., 2010). Yet, most studies assess interspecific variation and relationships with diet in adult specimens. Few analyses actually focus on their variation along the ontogeny, even though several of the studied functional attributes depend directly on the morphology, which itself changes along ontogenetic stages (Adams & Pedersen, 2000; Flores et al., 2018; Herrel & Gibb, 2006; Herring, 1993).

Morphological changes along the ontogeny are easily observed in marsupials, especially at early stages (Flores et al., 2003; Petrides, 1949). Unlike placentals, marsupials have a short gestation period and are born at an early anatomical development stage (Clark & Smith, 1993; Jurgelski, 1984; Smith & Keyte, 2020) in which selection on functional demands has prioritized the early development of the forelimbs and oral apparatus, used for crawling towards the teats and attaching to them (Gemmell et al., 2002). At birth, marsupial neonates have an ossified skull in the oral region and well‐organized tongue muscles that begin to mature before birth (Bennett & Goswami, 2013; Clark & Smith, 1993; Smith & Keyte, 2020). However, more striking morphological changes in the feeding apparatus occur during postnatal development, especially when weaned individuals develop features for mastication (Langenbach & Weijs, 1990; Stanchak et al., 2023). These changes generally include morphological changes in the head related to feeding habits and the development of specific dental occlusion patterns (Gorniak, 1985; Herring, 1985).

The mammalian masticatory apparatus consists mostly of the cranium, mandible, teeth, and masticatory muscles. The ontogeny of the cranial and mandible has been studied in the main groups of living marsupials (Flores et al., 2023), and among these, in Didelphidae, the largest radiation of marsupials outside Australasia (Voss & Jansa, 2021). Opossums are usually seen as morphologically unspecialized and of generalist feeding habits (Astúa & Guilhon, 2023; Lessa et al., 2023), yet represent almost a third of all marsupial species (Astúa et al., 2023).

The main ontogenetic changes in the skull of didelphids are associated with the masticatory apparatus, with most cranial measurements following an allometric growth pattern that follows the transition to adult performance, especially for the masticatory muscles and jaw strength (Flores et al., 2018). The cranial ontogenetic trajectories describe increasing space for the adductor muscles, suggesting an increase in the ability to produce stronger bite forces as individuals grow. This indicates that younger individuals face a greater challenge early on in development, when they are weaned and still have a smaller size, less‐developed muscles, a less robust skull and jaw, and incomplete dentition. One way to compensate for the bite force would be to improve the biomechanics of the apparatus early in ontogeny (Binder & Van Valkenburgh, 2000). The mechanical advantage, considering the mandible as a lever (Cleuren et al., 1995), can be estimated from the relationship between the lever arms of the muscles and the resistance (or bite) applied to the teeth (La Croix et al., 2011; Tanner et al., 2010). Thompson et al. (2003), however, showed that juveniles of the grey short‐tailed opossum Monodelphis domestica do not compensate for their weaker feeding system by improving the mechanical advantage of the adductor muscles. Although this study does not extend widely to other didelphid genera, ontogenetic variation in mechanical advantage has been inferred in ontogenetic studies of the skull, as related to rostrum length and mandible variables (Flores et al., 2023).

Masticatory muscle morphology has been rarely studied in didelphids (Diogo et al., 2016), except for Didelphis. Even though the cranium and mandible have been studied in detail in a wide range of taxa (Abdala et al., 2001; Astúa, 2010, 2015; Chemisquy et al., 2021; Flores et al., 2023), even providing inferences for muscle morphology (Astúa de Moraes et al., 2000; Silva‐Neto et al., 2023), Didelphis is the only genus with well‐documented anatomy of the masticatory muscles (Astúa & Guilhon, 2023; Coues, 1872; Delupi et al., 1997; Hiiemäe & Jenkins Jr, 1969; Minkoff et al., 1979; Turnbull, 1970). Its apparatus resembles that of carnivores with the dominance of the temporal muscle, followed by the masseter and pterygoid muscles, but it is considered to be generalized, as it presents features that are intermediate between other specialized groups (Turnbull, 1970). This morphology is consistent with its generalist feeding habit, even among didelphids (Lessa et al., 2023).

In addition, Diogo et al. (2016) showed that Didelphis retains a relatively more plesiomorphic muscle anatomy among therian mammals, as some earlier analyses had suggested for all opossums (Clark & Smith, 1993; Hiiemäe & Jenkins Jr, 1969; Radinsky, 1987; Weijs, 1994). Although analyses of other morphological systems or features have demonstrated that large opossums such as Didelphis are not the ideal models of a primitive condition (Amador & Giannini, 2016; Astúa, 2009; Chemisquy et al., 2021; Silva‐Neto et al., 2023), this genus has been routinely used as a model for early therian mammal masticatory morphology due to this conserved body morphology, to its size (which result in easier dissections) and to its availability throughout its range. As a result, Didelphis has offered a putative basic plan for studying morphological evolution (Fahn‐Lai et al., 2020; Radinsky, 1987), or more specifically masticatory performance (Delupi et al., 1997; Stilson et al., 2023; Thomason et al., 1990). This, however, is restricted to adult specimens.

Didelphis albiventris can reach up to 2500 g (Astúa, 2015), maintaining a generalized diet similar to other Didelphidae (vertebrates, fruit, and mainly invertebrates), and being one of the Didelphidae species with the best‐studied diet (Lessa et al., 2023). But while the post‐weaning ontogeny of the skull of Didelphis albiventris has been described, pointing to morphological changes along its growth that could be related to masticatory muscles and their function (Abdala et al., 2001), no proper analysis of the ontogeny of masticatory muscles has yet been done in this or any other opossum.

Thus, in this study, we describe the post‐weaning development of the masticatory muscles along the ontogeny in Didelphis albiventris, to provide a model for studies on masticatory muscles in Didelphidae and, ultimately, in earlier therian mammals. For that, we dissected specimens from an ontogenetic series (from pups to adults), recording changes in muscle origins, attachments, fiber orientation, and masses, along with functional variables. We expected such changes to reflect the changes in diet, from lactation to adult diet.

2. MATERIALS AND METHODS

2.1. Sample

We dissected the masticatory muscles of 25 specimens of Didelphis albiventris housed in the Mammal Collection of the Universidade Federal de Pernambuco (UFPE). We classified the studied specimens into four age classes based on the eruption of molars determined on the skulls, after the dissection: Class 1 (C1), before the full eruption of the first upper molar (M1) (12 specimens; <3.5 months old); Class 2 (C2), with deciduous upper third premolars (dP3) still present (4 specimens; 3.5–8 months old); Class 3 (C3), with permanent upper third premolars (P3), but lacking or with fourth upper molars (M4) incompletely erupted (1 specimen; 7.5–9.5 months old); Class 4 (C4), with the full adult dentition present (8 specimens; over 9 months old) (Figure. Molars were not considered fully erupted until they aligned in occlusion, even if they emerged from the dental alveoli (same criteria as used by Van Nievelt & Smith, 2005). The last three classes correspond to the juvenile, subadult, and adult age stages, respectively, used by Abdala et al. (2001), based on Regidor and Gorostiague (1990). The estimated age of the specimens was based on the onset and end of the molar eruption stages (Regidor & Gorostiague, 1990, 1996). All specimens in class C1 were still suckling. Most were collected along with the lactating mother, and the younger age class (3.5–4.5 months) established by Regidor and Gorostiague (1990) at weaning consisted of specimens with the two upper molars erupted, thus much older than our C1 specimens. Some adults may have reached up to 13 months old, as some of the molars showed some wear (Regidor & Gorostiague, 1990). Although sexual dimorphism has been reported for D. albiventris (Astúa, 2010, 2015), we did not make a distinction between sexes at these stages due to our sample sizes and because we expect that differences between age classes are likely to be more important than between sexes.

2.2. Dissection

The weight of the specimens ranged from 5.1 to 920 g, with cranial lengths ranging from 21.9 to 89 mm. The heads were frozen or fixed in 70% ethanol, with the mandibles near occlusion. Dissection was based on previous descriptions for the genus that included masticatory muscles (temporalis, masseter, pterygoideus medialis, and digastricus) and accessory muscles attached to the mandible (mylohyoideus and geniohyoideus) (Delupi et al., 1997; Hiiemäe & Jenkins Jr, 1969; Turnbull, 1970). Myologic nomenclature followed Diogo et al. (2016), and osteologic nomenclature followed Wible (2003). We inspected the crania and mandibles to determine muscular origin and insertion areas and fiber directions, and used these features to identify the different layers within muscular groups and compare them between classes. Dissections were conducted under a stereomicroscope and each step was photographed to map all fixation areas. All removed muscles were placed in 70% ethanol and weighed on a digital scale (0.01 g, Marte AD3300; or 0.001 g, Modus Pro‐50). Weights represent the muscle from one side or the average muscle on both sides, except for the mylohyoideus and geniohyoideus, which were removed and weighed bilaterally. Muscles sectioned or damaged for anatomical identification were not considered, resulting in different sample sizes depending on the muscle. Specimens in alcohol had their masses corrected by 40%, based on the percent change of mass obtained by Leonard et al. (2022) in a short period in ethanol from fresh muscle (but no formalin pre‐fixation was used in our sample). Only three specimens from the C1 class had their masses recorded.

2.3. Biomechanical model

We estimated the mechanical advantage for bite (or pitch, rotation of the jaw around the mediolateral axis that passes between the temporomandibular joints) of the main adductor muscles (m. temporalis and m. masseter) for the dissected specimens of classes C2–C4 using a bi‐dimensional lever model (Greaves, 2012). Specimens of the C1 class were not considered as they do not actively chew. We took the measurements on photographs of the cranium in lateral view with the mandible articulated. We calculated the mechanical advantage as the relation of the lever arms of the temporalis and masseter muscles and the lever arm of the reaction force at the canine and the first molar, representing the bite force (Brassard, Merlin, Guintard, Monchâtre‐Leroy, Barrat, Callou, et al., 2020). We computed the lever arms for the muscle forces as the perpendicular distance of the center point of the temporo‐mandibular joint (TMJ), an estimate of the rotation point for the pitch, to the line of action connecting the centroids of the origin and insertion areas in the cranium and the mandible. We established these areas based on the dissection photographs and marks on the bones. We computed the bite force lever arms as the perpendicular distance from the TMJ to the vector of the reaction force at the two bite points. The orientation of this force is particularly difficult to predict, but we considered an orientation of ca. 90° relative to the occlusion line passing through the incisives and the TMJ. We obtained all measurements using ImageJ 1.53 k (Schneider et al., 2012).

2.4. Analyses

We ran reduced major axes (RMA) regressions of the muscle weights on two size proxies (skull length and body weight) to assess scaling patterns along ontogeny. Regressions of mechanical advantages were run only with skull length. Muscle weights represented the individual weights of the mandibular adductor muscle groups (temporalis, masseter, and pterygoideus medialis), the overall weight of the adductor muscles (the sum of all adductor muscles individual weights), the weight of the digastricus and of the mylohyoideus and geniohyoideus combined. We measured the crania with a Mitutoyo digital caliper (0.01 mm), from the occipital condyles to the anteriormost point of the maxilla. Body weight, when available, was retrieved from the specimen tags or preparation sheets.

Before the analyses, all data were log‐transformed. The angular coefficient (b) describes the slope of the regression and was used to assess deviations from isometry. Weights were linearized to their cubic root when regressed against cranial length, to standardize isometry = 1 (Jungers & German, 2005; Mandarim‐de‐Lacerda, 2019). We tested for deviations from isometry using t‐tests in addition to confidence intervals of 99.8%. We used significance levels of 0.002 (after Bonferroni corrections for an α = 0.05), but we considered as moderate evidence of a deviation when p ranged from 0.002 to 0.05 (as in Flores et al., 2003). We also considered moderate evidence for confidence intervals of 95%. We ran regressions in PAST 4.03 (Hammer et al., 2001) and we used the linear model in R (R Core Team, 2021) in Rstudio (RStudio Team, 2020) to prepare the graphs using the ggplot function (Wickham, 2016) from Tidyverse.

3. RESULTS

3.1. Anatomical description

3.1.1. Masseter (m. masseter)

The masseter complex is divided into two portions, a superficial and a deep bundle, in all dissected classes. The superficial masseter (m. masseter pars superficialis) originated above the last upper molars, on the posterior end of the maxilla, and inserted on the ventral border of the angular process of the mandible (Ms; Figures 2, 3, and 4), with its fibers running mostly in a posteroventral direction. Its origin is more easily identified in older specimens (C3 and C4) through a tubercle on the maxilla, which is less obvious in younger specimens (C1 and C2). In pups (C1), particularly, it is attached through the masseteric fascia on the maxilla periosteum. The masseteric fascia covers both portions of the masseter, but it is thicker over the superficial masseter throughout its anterior extension on the maxilla and its posterior extension on the ear region. This clearly defined the superficial masseter in C1 and C2. The tendinous origin of the superficial masseter in C3 and C4 was located above the third upper molar or between the third and fourth upper molars, respectively, but some fibers reach more anteriorly, between the m. levator labii superiores and the m. buccinatorius. The superficial masseter also extends over the postero‐ventral border of the mandible in later ontogenetic stages (Ms; Figure 4). The additional fibers showed lower angles or were completely horizontally oriented (Ms; Figure 2), thus resulting in a greater diversity of fiber orientations. In a single C4 specimen, we observed an additional tendon closer to the mandible insertion, fixated to the mandibular horizontal ramus, below the m. buccinatorius and the line of the m3 (C4‐Ms; Figure 2), but fiber projections were seen in this region in other C4 specimens.

FIGURE 2.

Muscle locations in the age classes examined, in lateral views of specimens from classes C1 to C4 (Refer to text for age classes definitions). Muscles depicted: Temporalis, pars superficialis (Ts), and pars profunda (Tp); masseter, pars superficialis (Ms) and pars profunda (Mp); and digastricus (Dig).

FIGURE 3.

Muscles masseter pars superficialis (Ms), digastricus (Dig), mylohyoideus (Mh), and geniohyoideus (Gh), in ventral view of the cranium on the left, and pterygoideus medialis (Pm) on the right, in specimens from classes C1 and C4 (refer to text for age class definition). The digastricus and mylohyoideus are sectioned to show the relationships with the underlying muscles in the image. The geniohyoideus was only clearly distinct in C4 specimens. In C1 specimens, only the hyoglossus (Hg) and the genioglossus (Gg) were identified after the removal of the mylohyoideus, even though the fibers behind the symphysis were visualized.

FIGURE 4.

Fixation areas in the mandible of the dissected muscles. Muscles: Temporalis, pars superficialis (Ts), and pars profunda (Tp); masseter, pars superficialis (Ms) and pars profunda (Mp); digastricus (Dig), mylohyoideus (Mh), geniohyoideus (Gh), and pterygoideus medialis (Pm).

The deep masseter (m. masseter pars profunda) is partially covered by the superficial masseter, except when close to its origin on the zygomatic arch (Mp; Figure 2). Fibers for this portion originated from the medial surface of the zygomatic arch and inserted in the masseteric fossa, on the mandible (Mp; Figure 4). As in the superficial masseter, the fibers of the deep masseter are anteriorly oriented but differ in having a more vertical direction (Mp; Figure 2). Its position remained mostly constant across classes, except for the extension in the fossa towards the masseteric line, on the ventral limit (Mp; Figure 4). Although it approaches the ventral and posterior border of the mandible (from class C2 onwards), the muscular mass was mostly distributed closer to the origin.

3.1.2. Temporal (m. temporalis)

The temporal, just as the masseter, was divided into two parts, a superficial and a deep bundle. They have their origins posterior to the orbit and cover the lateral wall of the braincase along the ontogeny. The superficial temporal (m. temporalis pars superficialis) inserts on the lateral side coronoid process (Ts; Figures 2 and 4) and fills the space between the mandible, the deep masseter, and the upper medial surface of the zygomatic arch. In C1, although the fascia covers the cranium at the orbit level, the temporal mass is concentrated closer to its insertion, slightly upper than the position of the coronoid process, and extends anteriorly reaching behind the orbit and posteriorly reaching the posterior end of the zygomatic arch (C1‐Ts; Figure 2). Fiber orientation changes throughout the muscle extension, from posteriorly to anteriorly oriented as they approach the orbit. This anteriorly oriented component remains close to the insertion on posterior stages, while fibers on the braincase wall are more posteriorly oriented (Ts; Figure 2). In C2, the origin of the superficial temporal extended posteriorly on the braincase and anteriorly on the postorbital margin (C2‐Ts; Figure 2). The insertion on the mandible also extended inferiorly over the coronoid process. The fixation area of this portion on the braincase lateral wall and on the mandible keeps increasing in C3 and C4 (Ts; Figures 2 and 4). This portion becomes massive in older specimens, with its fixation area increasing and reaching dorsocaudally the nuchal and sagittal crests. The anterior extension above the orbit remains mostly constant from C2 onwards and was limited by the postorbital process in C3 and C4.

The deep temporal (m. temporalis pars profunda) is covered by the superficial portion and inserts on the coronoid process medially (Tp; Figure 4). It has its origin on the braincase wall posterior to the postorbital constriction, and the shape changes in the cranium in the early stage result in a particular situation. The lack of a postorbital constriction and the proximity between the braincase wall and the tip of the coronoid process in C1 leaves no room for it (C1, Figure 1). The deep masseter in these specimens was mostly located anteriorly and posteriorly to the coronoid process, along its borders (C1‐Tp; Figure 2). Its components, anterior and posterior, may reach one another below the tip of the coronoid process on the braincase wall, but its fibers insert on the mandible mostly through two muscular bundles, from each component, that join medially slightly below the dental line. From C2 onwards, the deep temporal covers medially the coronoid process and extends over the temporal fossa with the development of the postorbital constriction (C2‐Tp; Figures 2 and 4). As the superficial portion, the deep temporal also extends dorsally and posteriorly on the cranium until reaching the nuchal and sagittal crests in C4 and extends anteriorly below the border of the postorbital process. At this stage, fibers were mostly oriented posteriorly, although those that insert more rostrally on the coronoid process have a more vertical orientation. These fibers are more anteriorly oriented in age classes before C4 (Tp; Figure 2). The insertion of the deep temporal reaches the border of the coronoid process in C2 and extends under the border of the coronoid process in C3‐C4, from its base to its tip, covered by an aponeurosis (Tp; Figures 2 and 4).

FIGURE 1.

Ventral views of the cranium, lateral views of the cranium and mandible, and dorsal views of the cranium and mandible in occlusion, of representatives of the four age classes (C1–C4) used to separate the specimens. Refer to the text for class definitions. Adult crania (C3 and C4) have well‐developed sagittal and nuchal crests, longer rostra, and prominent postorbital constrictions. Younger skulls tend to have a shorter snout, with a rounded braincase that maintains close contact with the coronoid process of the mandible (arrow). Postorbital constriction is absent in these specimens.

3.1.3. Medial pterygoid (m. pterygoideus medialis)

The medial pterygoid muscle (also sometimes referred to as m. pterygoideus internus) has its origin extending posterolaterally onto the palatine, the pterygoid bone, and the alisphenoid, and it inserts on the medial side of the angular process in the mandible (Pm; Figures 3 and 4). Its origin on the cranium in C1 also reaches the ectotympanic. The space for the muscle increases in subsequent age classes with the increase of the pterygoid bone and the development of the postorbital constriction. The insertion of this muscle is located above the insertion of the superficial masseter, and below that of the deep temporal, and also increased along ontogenetic classes (Pm; Figure 4). In C1, the medial pterygoid inserted onto the angular process, and occupied a very small surface of the mandible. In subsequent classes, part of the muscle covers the horizontal ramus and reaches below the fourth lower molar in C4, with a strong tendon attached to the tip of the angular process. The muscle is anteriorly oriented, but increases in complexity as the specimens age. Fibers got more horizontal as they extended on the mandible along the ontogeny, while their trajectories were more restricted in C1.

3.1.4. Digastric (m. digastricus)

The digastric is a muscle composed of two bellies that has its origin in the basicranium and insertion on the mandible (Dig; Figures 2, 3, and 4). The posterior belly originates from the paracondylar process of the exoccipital and inserts onto the hyoid apparatus. The process is located just below the condyles, and is less distinct in the early stages, especially in C1 when it is closer to the tympanic process of the petrosal. It is small in C2 but increases along the development. This change affects the morphology of the posterior belly, which increases in volume, despite being similar to the anterior belly in C1. The posterior belly was even shorter than the anterior belly in C3 and C4. The division between the bellies is barely visible in C1 and C2, but it is possible to observe a slight constriction in C3 and the aponeurosis connecting the digastric to the hyoid apparatus in C4 (this connection was also tendinous in one of the specimens). The anterior belly originated in the hyoid apparatus and inserted medially into the mandible near the ventral border, anterior to the superficial masseter (Dig; Figures 3 and 4). In addition to its interaction with the superficial masseter towards its insertion, this belly partially covers the m. mylohyoideus, which is positioned just above it. This overlap is closer to the mandible in C1 and extends medially in later stages, but the contralateral bellies do not reach the midline. The anterior belly was wider and flatter than the posterior belly from C2 onwards. The fibers ran parallel in the two bellies.

3.1.5. Mylohyoid (m. mylohyoideus) and geniohyoid (m. geniohyoideus)

Along with the digastric, the mylohyoid and geniohyoid muscles constitute the floor of the oral cavity. The mylohyoid is a bilateral muscle that has its sides fused in the midline and covers the geniohyoid, located above (Figure 3). Unlike the other muscles previously presented, both the mylohyoid and the geniohyoid originate in the mandible and insert into the hyoid apparatus. The origin of the mylohyoideus extends anteriorly to the masseter attachment and above the digastric on the medial side of the mandible but does not reach the mandibular symphysis (Mh; Figure 4). Thus, it does not completely fill the intermandibular space, leaving the origin of the geniohyoid visible, positioned above, immediately behind the symphysis (Figures 3 and 4). The bilateral geniohyoids also fuse along the midline and, like the mylohyoids, do not have a fully visible raphe along their length. They are constituted by bulky bellies that insert into the hyoid apparatus, just above the mylohyoid (Gh; Figure 3), but we could not identify them in all specimens. The geniohyoids were only clearly distinguishable in C4, while in all previous stages, they were closely related to the mylohyoid, especially in C1. It is possible that the fact that the skull in C3 was conserved in alcohol influenced the misidentification and joint dissection of the two muscles. It was not, however, identified in C1 and C2 when dissecting fresh skulls. From C3 onwards, we observed an accumulation of fibers behind the symphysis, continuing up to the anterior limit of the mylohyoid. The two muscles have parallel fibers that project from the origin to the insertion (Figure 3).

3.2. Muscle masses

The muscle mass values are shown in Table 1. The proportion (% ± standard deviation) of the adductor muscles indicated a consistent prevalence of the temporal group: superficial temporal (C1: 21% ± 10.54; C2: 28.11% ± 14.12; C3: 49.37%; C4: 41.15% ± 3.53), deep temporal (C1: 32.51% ± 25.22; C2: 42.34% ± 28.31; C3: 17.02%; C4: 21.75% ± 6.22), superficial masseter (C1: 25.6% ± 15.74; C2: 17.73% ± 9.12; C3: 22.69%; C4: 21.65% ± 1.54), deep masseter (C1: 8.19% ± 1.62; C2: 6.31% ± 2.68; C3: 6.09%; C4: 10.33% ± 4.06), and medial pterygoid (C1: 12.71% ± 3.64; C2: 5.51% ± 3.78; C3: 4.83%; C4: 5.12% ± 41.77) (Figure 5). However, the higher proportion of the deep temporal in classes C1 and C2 was influenced by specimens DAM919, in C1, and DAM870, in C2. These specimens were the largest and smallest respectively in their class (based on cranial length). The deep temporal accounted for 61.54% in DAM919 and 75% in DAM870 of the total adductor mass, followed by the superficial temporal (10.99% and 12.5%), superficial masseter (8.79% and 7.5%), deep masseter (9.89% and 3.75%) and medial pterygoid (8.79% and 1.25%). When we removed specimen DAM870 from the C2 class (hereafter referred to as C2*), the distribution of the proportion of adductor muscles from C2 to C4 was more similar, with the superficial layer of the temporalis more prominent (Figure 5). The mass of the digastric and mylohyoid and geniohyoid combined compared to the unilateral total adductor mass was greater in C1; C1: 9.31% ± 3.55 and 49.24% ± 23.5, C2: 1.96% ± 0.62 and 7.72 ± 7.18, C3: 3.36% and 12.81%, C4: 3.11% ± 0.51 and 9.96 ± 3.38%.

TABLE 1.

Body weight (in g), cranial length (in mm), and muscle masses (in g, before correction) for the specimens examined in his study.

Specimen	Age Class	Sex	State	Body Weight	Skull Length	Ts	Tp	Ms	Mp	Pm	Total (adductors)	Dig	Mh + Gh
DAM845	C4	♀	Fresh	704.00	82.34	–	–	–	–	–	–	0.37	–
DAM866	C4	U	70% Ethanol	–	74.97	1.57	0.28	0.56	0.37	–	–	0.12	–
DAM891	C4	♂	Fresh	800.00	87.94	4.24 a	2.34	2.16 a	1.245 a	0.35	10.335	0.295 a	0.79
DAM900	C4	♀ (Lac)	70% Ethanol	920.00	88.97	2.84	1.20	1.79	1.37	0.46	7.66	0.22	0.79
DAM902	C4	♂	Fresh	720.00	86.19	3.48	2.92	1.73	0.69	0.28	9.10	0.26	0.81
DAM903	C4	♂	Fresh	760.00	87.13	5.00	3.15	2.84	1.00	0.55	12.54	0.33	0.69
DAM904	C4	♀ (Lac)	Fresh	600.00	75.55	2.64	0.96	1.28	0.51	0.37	5.76	0.20	0.72
DAM905	C4	♀ (Lac)	Fresh	625.00	84.49	2.47	1.01	1.20	0.42	0.41	5.51	0.22	0.82
DAM873	C3	♂	70% Ethanol	530.00	75.51	2.35	0.81	1.08	0.29	0.23	4.76	0.16	0.61
DAM869	C2	U	70% Ethanol	300.00	65.31	0.60	–	0.44	0.20	0.08	–	0.07	0.12
DAM870	C2	U	70% Ethanol	70.00	45.65	0.10	0.60	0.06	0.03	0.01	0.800	0.01	0.02
DAM899	C2	U	Fresh	324.00	–	1.18	0.73	0.61	0.18	0.25	2.95	0.07	0.14
DAM908	C2	♂	Fresh	113.00	49.20	0.14	0.12	0.11	0.04	0.03	0.44	0.01	0.07
DAM872	C1	U	70% Ethanol	22.00	32.26	–	–	–	–	–	–	–	–
DAM882	C1	U	Fresh	6.500	–	–	–	–	–	–	–	–	–
DAM884	C1	U	70% Ethanol	5.100	22.15	–	–	–	–	–	–	–	–
DAM885	C1	U	70% Ethanol	6.00	21.92	–	–	–	–	–	–	–	–
DAM887	C1	U	70% Ethanol	10.00	24.39	–	–	–	–	–	–	–	–
DAM888	C1	U	70% Ethanol	8.00	22.68	–	–	–	–	–	–	–	–
DAM889	C1	U	70% Ethanol	6.25	22.48	–	–	–	–	–	–	–	–
DAM890	C1	♂	70% Ethanol	9.25	–	–	–	–	–	–	–	–	–
DAM892	C1	U	Fresh	10.50	29.49	–	–	–	–	–	–	–	–
DAM916	C1	U	Fresh	15.5	26.45	0.004	0.002	0.0035 a	0.001 a	0.002	0.0125	0.001	0.006
DAM917	C1	U	Fresh	16	26.08	0.003	0.003	0.006 a	0.001	0.002	0.015	0.002	0.011
DAM919	C1	U	Fresh	14.5	30.37	0.005	0.028	0.004 a	0.0045 a	0.004	0.0455	0.003	0.012

Abbreviations: Dig, Digastricus; Lac, Lactating; Mh + Gh, Mylohyoideus and geniohyoideus; Mp, Masseter pars profunda; Ms, Masseter pars superficialis; Pm, Pterygoideus medialis; Tp, Temporalis pars profunda; Ts, Temporalis pars superficialis; U, Undetermined.

^a The mass was represented by the average of the muscle taken from both sides of the head.

FIGURE 5.

Average percentage mass of the mandibular adductor muscles (m. temporalis, m. masseter, and m. pterygoideus medialis) of specimens from classes C1 to C4 (see text for age class definition). In each graph, the proportion of muscles is compared with the previous stage (C1–2, C2–C3, C3–C4). C2* represents the average percentage mass of the class C2 without specimen DAM870.

3.3. Allometry

Muscle masses were strongly correlated with the two size indicators (p < 0.001) (Table 2). Both skull size and body mass were good predictors of muscle mass. The consistency of the results in the models using skull length and body mass (a variable more independent of the masticatory apparatus) supports a pattern of muscle growth essentially by positive allometry, except for the mass of the deep temporal and mylohyoid and geniohyoid combined, which were isometric with body mass (Figure 6; Table 2). There was moderate evidence for positive allometry for both the masses of the deep temporal, and mylohyoid and geniohyoid combined (Table 2).

TABLE 2.

Reduced major axis (RMA) regressions statistics of muscle mass on skull length and body mass and mechanical advantage on skull length.

	Cranial length					Body mass
	Slope	r	R 2	95% C.I.	99.8% C.I.	Slope	r	R 2	95% C.I.	99.8% C.I.
Muscles mass
Temporalis pars superficialis	2.025	0.994	0.989	1.891–2.159	1.783–2.267	1.760	0.993	0.987	1.633–1.886	1.531–1.988
Temporalis pars profunda	1.747	0.939	0.881	1.347–2.147	1.015–2.478	1.537	0.919	0.844	1.135–1.940	0.801–2.273
Masseter pars superficialis	1.795	0.991	0.982	1.646–1.944	1.526–2.064	1.568	0.996	0.991	1.476–1.659	1.403–1.733
Masseter pars profunda	1.867	0.991	0.982	1.711–2.023	1.586–2.149	1.617	0.982	0.964	1.423–1.810	1.268–1.966
Pterygoideus medialis	1.548	0.992	0.983	1.415–1.681	1.305–1.791	1.358	0.990	0.979	1.235–1.480	1.136–1.579
Total (adductors)	1.792	0.987	0.974	1.589–1.994	1.415–2.169	1.559	0.976	0.953	1.334–1.784	1.148–1.971
Digastricus	1.548	0.988	0.976	1.405–1.691	1.293–1.803	1.339	0.985	0.971	1.202–1.476	1.094–1.584
Mylo‐ and geniohyoideus	1.385	0.981	0.962	1.206–1.563	1.058–1.711	1.200	0.980	0.961	1.052–1.349	0.933–1.468
Mechanical advantage (molar 1)
Temporalis pars superficialis	0.521*	0.581*	0.338	0.222–0.820	(−0.034)–1.077
Temporalis pars profunda	0.379*	0.189	0.036	0.117–0.641	(−0.109)‐0.867
Masseter pars superficialis	−0.615*	−0.569	0.324	(−0.972)–(−0.259)	(−1.278)‐0.047
Masseter pars profunda	−0.437	−0.781*	0.611	(−0.629)‐(−0.245)	(−0.794)–(−0.080)
Mechanical advantage (canine)
Temporalis pars superficialis	0.737	0.821	0.673	0.440–1.033	0.185–1.288
Temporalis pars profunda	0.535	0.698*	0.487	0.265–0.805	0.033–1.037
Masseter pars superficialis	−0.539*	−0.091	0.008	(−0.917)–(−0.161)	(−1.242)–0.164
Masseter pars profunda	−0.299*	−0.133	0.018	(−0.508)–(−0.090)	(−0.688)–0.089

Note: All regressions were performed with log‐transformed values. Mass was linearized to ^1/3 before analysis with skull length. Values in bold highlight significant p (slope and correlation, α = 0.002) and the confidence interval (C.I.; 95% and 99.8%) whenever the obtained was outside the interval expected for isometry (= 1).

^* p ranged between 0.002 and 0.05.

FIGURE 6.

Regressions of muscle mass on skull length and body mass. Muscles: Temporalis, pars superficialis (Ts) and pars profunda (Tp); masseter, pars superficialis (Ms) and pars profunda (Mp), pterygoideus medialis (Pm), digastricus (Dig), mylohyoideus (Mh) and geniohyoideus (Gh). Separated graphs of the regressions by muscle in Figure S1. See text for age classes (C1–C4) definitions.

The relationships between the mechanical advantage of the main adductors and skull length were not evident for all the muscles (molar 1: deep temporal, superficial masseter; canine: superficial masseter, deep masseter, p > 0.05) (Table 2). Only the mechanical advantage of the superficial temporal at the canine was significantly correlated with skull length (p = 0.001), while the superficial temporal and the deep masseter for molar 1, and the deep temporal for the canine showed moderate evidence (0.05 > p > 0.002). All of these grew with isometry, except for the deep masseter for the molar, which was negative allometry (Figure 7; Table 2). However, only superficial temporal growth remained isometric when the 95% confidence interval was considered (Table 2).

FIGURE 7.

Regressions of the mechanical advantage of the temporal and masseter on skull length, for biting on the canine (c) and first molar (m1). Muscles: Temporalis, pars superficialis (Ts) and pars profunda (Tp); masseter, pars superficialis (Ms) and pars profunda (Mp). See text for age classes (C2–C4) definitions.

4. DISCUSSION

Although most myological studies of the masticatory apparatus of didelphid marsupials focused on the genus Didelphis (Coues, 1872; Delupi et al., 1997; Diogo et al., 2016; Hiiemäe & Jenkins Jr, 1969; Minkoff et al., 1979; Turnbull, 1970), and aside from Smith's (1994) pioneering work on the early development of craniofacial musculature in Monodelphis domestica, this is the first study to qualitatively and quantitatively assess muscle development throughout postnatal ontogeny in Didelphidae, and possibly in any marsupial, including the transition from suckling to mastication. Overall, the main differences found were between suckling (C1) and weaned (C2–C4) specimens, although some changes still occurred from juveniles to adults. Functionally and ecologically speaking, though, C2–C4 specimens can all be considered independently foraging and living individuals, as breeding occurs as early as in our C2 specimens (Astúa & Geise, 2006). Thus, the main differences between unweaned and weaned specimens can be initially related to the onset of mastication, and to a lesser extent, to size‐related changes in growing weaned age classes. In general, all adductor muscles grew with positive allometry, and increased their fixation areas through, in part, the development of specific regions of the cranium and mandible. Their lines of action also increase and diversify along ontogeny. These changes can be related to the transition from functional requirements for fixation during lactation, to adduction and mastication movements after weaning.

In general, our description of the older (C3 and C4) specimens was similar to previous descriptions for the genus. Throughout ontogeny, muscle anatomy varied mainly on the extension of the areas of attachment to the cranium and mandible throughout development, especially the adductor musculature (Figures 2, 3 and 4), but the complexity (i.e., number of muscles; Diogo et al., 2016) remained the same. Thus, we did not observe any layers other than the superficial and deep layers of the temporal and masseter, contrarily from Turnbull's (1970) description of D. marsupialis, who recognized the m. zygomaticomandibularis in that species, as did Deguchi et al. (2001). The differences observed between D. marsupialis and D. albiventris may be due to interspecific variations or between dissections. However, Turnbull (1970) stated that the m. zygomaticomandibularis originates on the medial surface of the zygomatic arch, with a more vertical orientation of the fibers than the deep masseter, but he also mentioned that the muscle was not clearly distinct from the deeper layers of the masseter, and closely united to the temporal (see also Hiiemäe & Jenkins Jr, 1969). The latter observation matches the description made by Diogo et al. (2016), who did not identify the m. zygomaticomandibularis as an individual muscle in Didelphis virginiana. They described it as a bundle of fibers joined to the masseter and partially to the temporal, which we also found. Finally, the absence of the m. zygomaticomandibularis in D. albiventris is corroborated by Delupi et al. (1997), who did not record this muscle as part of the cranial musculature in this species.

Although distinguishing the m. zygomaticomandibularis from the other masseter layers may be arbitrary and artificial (Hiiemäe & Jenkins Jr, 1969), the general change in fiber orientation (closer to the masseter rather than to the rest of the temporal) helps the distinction, as it could suggest a distinct action between the parts (Hiiemäe & Houston, 1971). However, the continuity of the deep masseter fibers with part of the insertion of the superficial temporal (see also Diogo et al., 2016; Hiiemäe & Jenkins Jr, 1969; Turnbull, 1970) indicates a functional chain of this adductor “block” (Dawson et al., 2014). Based on this and the uniformity observed in stage C2, we considered a second line of action for the superficial temporal, which we grouped with that of the deep masseter (Figure 8). This grouping is justified by the proximity between the lines of action, the fiber orientations, and the position between the adductor parts. Under this view, our description agrees with the functional description of the adductor musculature proposed by Hiiemäe and Jenkins Jr (1969), who separated it into three functional parts: (i) external adductor, which originates from the temporal fascia, zygomatic arch, and masseteric fascia, and inserts on the lateral surface of the coronoid process; (ii) internal adductor, which originates from the lateral wall of the braincase, anterior to the temporomandibular junction, and inserts on the medial surface of the coronoid process; and (iii) posterior adductor, with fibers that originate from the posterosuperior region of the temporal fossa and parts of the sagittal and nuchal crests, to insert on the posterior edge of the coronoid process. Despite some differences, our results converge with this general description. We also recognized three adductor parts in D. albiventris, represented by the deep masseter and the superficial temporal fibers which run from the zygomatic arch to insert laterally into the mandible (external adductor), by the deep temporal (internal adductor) and by the fibers of the superficial temporal which originate in the skull and converge towards the top of the coronoid process (posterior adductor).

FIGURE 8.

Estimated lines of action of the superficial (red) and deep (orange) temporals, and superficial (dark blue) and deep masseter (light blue) in lateral view, and superficial masseter (dark blue) and medial pterygoid (green) in ventral view of representatives of stages C1 to C4. We considered a second line of action for the superficial temporal since C2, based on the change in fiber orientation that suggests a distinct action between the parts of the muscle. See text for age class definitions. Muscles: Temporalis, pars superficialis (Ts) and pars profunda (Tp); masseter, pars superficialis (Ms) and pars profunda (Mp); and pterygoideus medialis (Pm).

When we analyze C1 specimens under this view (based on older specimens), the posterior component of the line of action of the superficial temporal is limited, and the deep temporal divides its action into two components, one more anterior and the other more posterior (Figure 8). As such, adduction and gape seem to be limited during lactation, based on the mentioned structure of the temporal, combined with the superficial masseter, represented by a thin layer of muscle, and the medial pterygoid, which apparently limits abduction of the mandible (Figures 2, 3 and 8). The reduction in mandibular movement at this stage is not surprising given that the tongue performs the main activity in the breast milk pumping apparatus (Filan, 1991; Herring, 1985), while the masticatory muscles act later in ontogeny (Brassard, Merlin, Guintard, Monchâtre‐Leroy, Barrat, Bausmayer, et al., 2020; Campbell‐Malone et al., 2011; German & Crompton, 1996, 2000). Minimal movements of the mandible during suckling have been reported for D. virginiana, and tongue movement is mainly dorsal/ventral at this stage (German & Crompton, 1996). The floor of the oral cavity also indicates this movement, and thus its muscles have been associated with suckling (German & Crompton, 1996; Langenbach & Weijs, 1990). The mylohyoid and geniohyoid take part in the oral cavity floor but grow with isometry or positive allometry rather than negative allometry, as would be expected for a lower performance of these muscles at latter ontogenetic stages, in which pups would gain some level of independence from their mothers, shifting from suckling to drinking (German & Crompton, 1996, 2000). However, even though the digastric, the mylohyoid, and the geniohyoid are grouped and maintain a common function during swallowing, and were proportionally larger compared to the adductor mass in C1, the mylohyoid is likely more effective for moving the oral cavity floor and elevating the tongue (Hiiemäe & Jenkins Jr, 1969). The digastric and geniohyoid, on the other hand, are involved in mandibular opening (Weijs et al., 1987), which is key for predation or consumption of larger items that can meet the energy requirements of the larger size of adult individuals. Thus, the growth pattern of the mass of the mylohyoid plus geniohyoid may be due to the geniohyoid, which was only clearly evident in C4 individuals. The mylohyoid may show a different growth pattern, closer to negative allometry, but the muscles were not removed individually from all the specimens to allow for a test of this hypothesis.

In C2 specimens, the function and strength of the adduction are initially improved by the development of the deep temporal (for internal adduction) and then by the superficial temporal (for posterior adduction). The superficial temporal remained dominant in C3 and C4 specimens. Throughout these stages, additionally to suspending the mandible, the masticatory apparatus increases its ability to adduct against an anterior resistance, as the posterior component of the temporalis fibers becomes more evident. The production of posterior force is a need observed in carnivorous mammals (Greaves, 1991) and the increase of this property along development is important for obtaining larger and harder prey. On the other hand, the superficial masseter includes more anterior, even horizontal, fibers throughout growth. These fibers oppose posterior adduction and can act simultaneously with the contralateral medial pterygoid in the mandible (Hiiemäe & Jenkins Jr, 1969).

An additional major change between unweaned (C1) and weaned (C2–C4) specimens occurs on the medial pterygoid (Figures 3 and 4) and is likely related to the onset of mastication itself. Mastication involves not only the abduction‐adduction of the mandibles but also complex mediolateral movements related to the action of the tribosphenic molars (Bhullar et al., 2019; Crompton & Hiiemäe, 1970; Grossnickle, 2017; Schultz & Martin, 2014; Stilson et al., 2023). Yaw (rotation of the whole mandible around a dorsoventrally oriented axis) and roll (rotation around the long axis rotation of the hemimandibles, allowed by an unfused symphysis) were recently documented in Didelphis by Stilson et al. (2023), but had already been recorded by Crompton and Hiiemäe (1970) in their studies on mastication in Virginia opossums. The importance of the roll not only to bring molars to the proper orientation but also for a specific transverse grinding motion was reported by Bhullar et al. (2019) in a detailed study of the mastication in the didelphid Monodelphis domestica, with similar movements later found in Didelphis virginiana (Stilson et al., 2023). Eversion and inversion movements in hemimandibles occur during the chewing cycle, due to the actions of the medial pterygoid on one side and the superficial masseter on the other (Bhullar et al., 2019). Thus, their action, unnecessary during the simple occlusion of the suckling pups becomes essential for the more complex occlusion in the mastication in weaned specimens, which is reflected in their development in classes C2 onwards. Additionally, Grossnickle (2020) argued that hemimandibles roll movements are facilitated by the inward inflection of the angular process that provides improved mechanical advantage for the medial pterygoid. Although the mechanical advantage for mediolateral movements of the mandible remains to be investigated throughout the ontogeny of D. albiventris, the variation in jaw movements is also consistent with the eruption of the molars, the increase in the volume of this muscle and the increased insertion area in the mandible, especially when suckling and weaned specimens are compared (Figures 3 and 4).

The action line and fiber orientations of the medial pterygoid tend to be more anteriorly directed as specimens grow, just as those of the masseter, but they reach a less vertical direction to the frontal plane as it expands into the mandible (Figures 4 and 8). This diversity in fiber orientation may contribute to the diversity of jaw movement in the different phases of mastication in Didelphis (Crompton & Hiiemäe, 1970). The fibers of the medial pterygoid could pull the lower border of the mandible anteriorly and medially, inducing a rotation in the axis of the hemimandible itself rather than the entire structure (Hiiemäe & Jenkins Jr, 1969). This medial component is reduced in suckling young (C1) (Figure 3). The fibers tend to be more vertically aligned in the inflexed angular process, which can favor muscle extension during depression of the mandible (Vinyard & Taylor, 2010). However, the proximity between the origin of the medial pterygoid and its insertion on the angular process at C1 reduces its line of action. The muscle at this stage seems to be acting more as a stabilizer of the mandible, preventing its abduction. It is possible that the reduction in jaw movements in C1 individuals, helped by the medial pterygoid, is not just a limitation of the adductor musculature, but rather a necessity to avoid possible damage to the cranium caused by muscle adduction, because, unlike the oral apparatus which is more ossified at birth, the ossification of the braincase occurs later in Didelphidae and in all marsupials (Bennett & Goswami, 2013; Clark & Smith, 1993; Smith & Keyte, 2020). This could represent a functional explanation for the presence of the inflection of the angular process typical of most marsupials, reducing the line of action of the medial pterygoid, at least at the beginning of development. Although tempting, this hypothesis still requires further study, especially including developmental studies of the jaw joint just after birth and comparing taxa lacking an inflected angular process (e.g., Caluromys), but ontogeny may be key to explaining this feature (Sánchez‐Villagra & Smith, 1997).

The allometry of muscle size is congruent with the cranial growth patterns of Didelphis albiventris (Abdala et al., 2001). The physical space for the temporal in the cranium (which is lacking in C1) results from the negative allometry of the braincase combined with the isometry of the zygomatic arch. The positive allometry of the coronoid process through growth also increases the surface for muscle insertion in the mandible. This pattern allows for the gradual increase of the temporal relative to the skull length throughout development but occurs differently in the superficial and deep layers with the body mass. Growth of the deep temporal was isometric, while that of the superficial temporal was allometric with body mass. The positive allometry of the latter is also likely related to the development of the sagittal crest, increasing the surface for the origin of the muscle in the cranium throughout ontogeny, while the deep layer depends more on the postorbital constriction. Although there is moderate evidence for a positive allometry of the deep temporal, an isometric growth suggests the importance of the muscle early in weaning (C2; Figure 5) regardless of skull size, as well as mylohyoid and geniohyoid. In contrast, the rest of the adductors present a positive allometric pattern consistent with body mass and skull length. This positive trend indicates that adult individuals have relatively larger muscles and probably a relatively stronger bite force, as muscle strength (indicated by physiological cross‐sectional area; PCSA) scales with muscle mass and fascicle length tend to remain the same or decrease with increasing size (Abreu et al., 2023).

The bite force is an important feature of the masticatory apparatus, as it relates to the array of food items accessible for consumption by an individual (Dickinson et al., 2023). Stronger bite forces in adults ensure access to harder items. Cáceres (2002) detected differences in natural diet composition between age classes of D. albiventris. However, he was unable to find significant differences between them (Cáceres, 2002). This indicates that young D. albiventris have a similar diet to adults, but have relatively smaller muscles and adductor mass. As a result, a greater mechanical advantage could be expected in these younger individuals, to compensate for the weaker masticatory apparatus (Binder & Van Valkenburgh, 2000). However, when we found some evidence of a correlation between mechanical advantage for pitch and skull length, they indicated only moderate evidence of this expectation, and tended towards isometry (Table 2). Previous studies (La Croix et al., 2011; Tanner et al., 2010; Thompson et al., 2003) also found no negative allometry, but the isometry contrasts with the positive allometry observed in these studies. Even though isometric growth reduces the difference in the advantage of the adult masticatory apparatus compared to juveniles in Didelphis, compared to what could be than expected by positive allometry, the difference with those studies may be due to the simpler approach to muscle strength, without considering the orientation of the muscles as we used here.

The isometric growth of the mechanical advantage supports the idea that the allometric increase in muscle size is more important than the mechanics of the mandible for the production of bite force (La Croix et al., 2011; Law et al., 2016; Stanchak et al., 2023). Adults seem to improve efficiency through behavior, when they position food on the posterior teeth, with higher mechanical advantage. The continuous and allometric growth of the length of the mandible in D. albiventris (Abdala et al., 2001; Flores et al., 2003, 2010, 2018, 2023) may favor gape, which is proportional to the distance of any point along the mandible from the condyle (Kiltie, 1984) and speed for larger and more agile items, such as birds and other mammals (Cáceres, 2002). In addition, an increased gape allows the use of more posterior teeth for larger items as well. On the other hand, gape is limited in younger specimens (Singleton, 2015). These may need to open their mouths wider or bite down more anteriorly (Binder & Van Valkenburgh, 2000), as would be allowed by the functional molariform deciduous premolar (Van Nievelt & Smith, 2005). Adduction of the temporal (dominant since C2*; Figure 5) can help in either case, as it can produce more force in a larger mandible opening (Herrel et al., 2008; Law et al., 2016; Thexton & Hiiemae, 1975). However, from C2* onwards there are no marked changes in the organization of the masticatory apparatus (Figure 5). The configuration of the mandibular muscles matures early and grows larger but without adjustments to the overall growth pattern (Figure 6). While this seems to provide some performance improvement in the younger classes, it may also be because Didelphis has been described as less specialized than other mammals (La Croix et al., 2011; Turnbull, 1970). This raises the question of whether didelphids with a less generalized masticatory system are driven by a pattern of skull shape growth that occurs even earlier, resulting in different allometric patterns of the mandibular muscles, as Lutreolina, for instance, shows a more developed adductor mass and the anticipation of skull features compared to D. albiventris (Delupi et al., 1997; Flores et al., 2003). This only emphasizes the need for additional descriptions of the post‐weaning ontogeny of the masticatory muscles in Didelphidae, especially in those taxa with more distinct cranial shapes or dietary preferences.

AUTHOR CONTRIBUTIONS

JAFHA: Data acquisition, data analysis, drafting of the manuscript, critical revision of the manuscript, and approval of the article. DA: Study design, data analysis, critical revision of the manuscript, approval of the article, funding.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

Supporting information

Figure S1.

Abreu, J.A.F.H. & Astúa, D. (2024) Ontogeny of the masticatory muscles in the opossum Didelphis albiventris (Marsupialia, Didelphimorphia, Didelphidae). Journal of Anatomy, 245, 625–642. Available from: 10.1111/joa.14109

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.