Skip to main content
NCBI home page
Integrative and Comparative Biology logoLink to Integrative and Comparative Biology
. 2011 Jun 29;51(2):247–259. doi: 10.1093/icb/icr068

A Preliminary Analysis of Correlated Evolution in Mammalian Chewing Motor Patterns

Susan H Williams *,1, Christopher J Vinyard †, Christine E Wall ‡, Alison H Doherty †, Alfred W Crompton §, William L Hylander ‡
PMCID: PMC3135829  PMID: 21719433

Abstract

Descriptive and quantitative analyses of electromyograms (EMG) from the jaw adductors during feeding in mammals have demonstrated both similarities and differences among species in chewing motor patterns. These observations have led to a number of hypotheses of the evolution of motor patterns, the most comprehensive of which was proposed by Weijs in 1994. Since then, new data have been collected and additional hypotheses for the evolution of motor patterns have been proposed. Here, we take advantage of these new data and a well-resolved species-level phylogeny for mammals to test for the correlated evolution of specific components of mammalian chewing motor patterns. We focus on the evolution of the coordination of working-side (WS) and balancing-side (BS) jaw adductors (i.e., Weijs’ Triplets I and II), the evolution of WS and BS muscle recruitment levels, and the evolution of asynchrony between pairs of muscles. We converted existing chewing EMG data into binary traits to incorporate as much data as possible and facilitate robust phylogenetic analyses. We then tested hypotheses of correlated evolution of these traits across our phylogeny using a maximum likelihood method and the Bayesian Markov Chain Monte Carlo method. Both sets of analyses yielded similar results highlighting the evolutionary changes that have occurred across mammals in chewing motor patterns. We find support for the correlated evolution of (1) Triplets I and II, (2) BS deep masseter asynchrony and Triplets I and II, (3) a relative delay in the activity of the BS deep masseter and a decrease in the ratio of WS to BS muscle recruitment levels, and (4) a relative delay in the activity of the BS deep masseter and a delay in the activity of the BS posterior temporalis. In contrast, changes in relative WS and BS activity levels across mammals are not correlated with Triplets I and II. Results from this work can be integrated with dietary and morphological data to better understand how feeding and the masticatory apparatus have evolved across mammals in the context of new masticatory demands.

Introduction

Nearly all studies aimed at understanding the motor control of feeding in vertebrates emphasize the inherent flexibility in jaw-muscle activity patterns within individuals during the acquisition and processing of food. As a potentially significant portion of the total variation in muscle activity, this flexibility has been defined as modulation (Deban et al. 2001), and is thought to reflect active adjustments by individuals in response to changes in the size, shape, and texture of food, as well as active responses to the environmental or behavioral context of food processing. Regardless of this flexibility, researchers studying the motor control of feeding have also stressed that there are consistently identifiable patterns of muscular contraction that characterize feeding behaviors both across individuals and among species. These muscle activity patterns, or motor patterns, are characterized by consistent order, duration and/or magnitude of muscle activation during specific feeding tasks (Wainwright and Friel 2001). Moreover, during feeding, different motor patterns have been functionally linked to differences in musculoskeletal morphology which in turn is often intimately tied to specific dietary niches and/or feeding behaviors.

The functional links among motor pattern, morphology and diet have been of particular interest to researchers studying mastication in mammals. Minimally, mastication emphasizes the reduction of food using the postcanine dentition. However, in most mammals mastication also occurs unilaterally with precise occlusion of the postcanine teeth and involves both vertical and transverse motion of the lower jaw during the power stroke (see Hiiemae 1976). To facilitate precise occlusion, bite force production and three-dimensional movements of the jaw during fast-closing and the power stroke, the jaw adductors are usually activated bilaterally but not simultaneously, thereby producing a complex motor pattern.

In one of the first comparisons of chewing motor patterns across mammals, Hiiemae (1978, p. 390) argued for the conservation of chewing motor patterns based on existing EMG data, stating that “It is clear that the pattern of EMG activity in the major jaw muscles is broadly similar in all mammals so far studied despite difference in the profile of movement and the structure of the jaw apparatus.” She further argued that the differences in jaw movements observed across mammals are dictated by the morphology of the masticatory apparatus, including the anatomy and architecture of the jaw muscles. Several years later, Smith (1994) critiqued broad-brush hypotheses for the conservation of feeding motor patterns, arguing that we need more rigorous and quantitative tests of the EMG data.

At about the same time, however, Weijs (1994) published his seminal work on the evolution of chewing motor patterns, in which he compared all of the existing EMG data collected during chewing in mammals. His stated goal was to “show that similar motor patterns evolved from a basic primitive motor pattern in different independent lineages in response to similar demands on the masticatory system” (Weijs 1994, p. 284). Thus, he argued that motor patterns evolve and that we should expect to see convergence in motor patterns related to diet. Although only qualitative, his interpretation of the EMG data identified several derived motor patterns linked to diets that evolved from a primitive mammalian pattern. More importantly, he identified groups of jaw adductors that contract together and coordinate jaw movements and forces, and hypothesized that changes in timing of these muscle groups results in the major kinematic differences we see across mammals. He called these muscle groups (1) Vertical Closers, consisting of the working-side (WS; i.e., chewing side) and balancing-side (BS; i.e., non-chewing-side) deep masseter, (2) Triplet I, consisting of the BS superficial masseter and medial pterygoid and WS posterior temporalis, and (3) Triplet II, consisting of the WS superficial masseter and medial pterygoid and BS posterior temporalis. Although Weijs (1994) was not the first to identify the coordination in activity among groups of jaw muscles (see Herring 1976, 1985, 1992, 1993), his comprehensive treatment of the EMG data has made a lasting contribution to an understanding of mammalian jaw-adductor function because of its mammal-wide comparative perspective.

Inspired by Weijs’ observations, several quantitative studies of chewing motor patterns in mammals have been published since 1994. Some of these studies have directly tested hypotheses of the evolution of motor patterns, identifying quantitative differences between closely-related species that have qualitatively similar chewing kinematics (e.g., Hylander et al. 2005; Vinyard et al. 2005; Williams et al. 2007; Crompton et al. 2008a). For example, Crompton et al. (2010; see also Crompton 2011) has demonstrated in a number of marsupial herbivores that the jaw muscles generally tend to fire in two groups, but that the constituent muscles in these groups differ among species and from Weijs’ Triplet hypothesis [Because of the variability in muscles within each group, Crompton et al. (2010; Crompton 2011) use the terms Group I and II to avoid confusion with Triplet I and II when describing jaw-adductor activity.]. Moreover, new jaw-adductor activity patterns have been identified that may have evolved in concert with changes in the morphology of the jaws and teeth (e.g., Hylander et al. 2004, 2005 and references therein; Crompton et al. 2008b, 2010; Vinyard et al. 2011). In addition to new scientific contributions, continued investigations on the motor control of mammalian chewing have significantly increased available EMG data, particularly from primates, ungulates and marsupials.

Here, we investigate the correlated evolution of chewing motor patterns across mammals. We test hypotheses for the co-evolution of specific components of the mammalian chewing motor pattern, focusing on the coordination of WS and BS jaw adductors (i.e., Weijs’ Triplets I and II), the evolution of relative WS and BS muscle recruitment levels, and the evolution of asynchrony between muscle pairs (Table 1). For example, using Weijs’ (1994) Triplet model, we test for the correlated evolution of Vertical Closer timing and the presence or absence of Triplets I and II across mammals. Our focus here on the correlated evolution of activity among the chewing muscles is complemented by the work on the homology of jaw muscles (Druzinsky et al. 2011) and the correlated evolution of jaw-muscle activity and masticatory morphology across mammals (Vinyard et al. 2011). Although these analyses are exploratory, they are timely because of new data collected since Weijs’ article (1994) and because of the improvements in the species-level phylogeny for Mammalia (Bininda-Emonds et al. 2007). Additionally, new efforts by the Mammalian Feeding Working Group, funded by the National Evolutionary Synthesis Center, to archive existing data in the Feeding Experiment End-User Database (FEED) and conduct synthetic analyses of these data have inspired new interest in the evolution of mammalian chewing motor patterns (Wall et al. 2011). Collectively, these advances justify this first attempt to bridge a gap noted a decade ago by Hiiemae (2000, p. 412), when she lamented that “possibly as a consequence of uncertainty in the higher level-relationships of mammals, virtually no study has attempted an overarching, phylogenetic analysis of feeding system evolution in mammals.”

Table 1.

Hypotheses for the correlated evolution of chewing motor patterns in mammals

Hypothesis 1 The evolution of Triplet I is correlated with the evolution of Triplet II.
Hypothesis 2A–C The evolution of Triplet I is correlated with changes in the relative amount of WS and BS muscle activity from the superficial masseter (2A), medial pterygoid (2B), and posterior temporalis (2C).
Hypothesis 3A–C The evolution of Triplet II is correlated with changes in the relative amount of WS and BS muscle activity from the superficial masseter (3A), medial pterygoid (3B), and posterior temporalis (3C).
Hypothesis 4A and B The evolution of asynchrony of WS and BS deep masseter activity is correlated with the evolution of Triplet I (4A) and Triplet II (4B).
Hypothesis 5 Shifts in the relative timing of the BS deep masseter are correlated with changes in the relative amount of WS and BS deep masseter activity.
Hypothesis 6 Shifts in the relative timing of the BS deep masseter are correlated with shifts in relative timing of the BS posterior temporalis.
Open in a new tab

Note. WS and BS refer to working-side and balancing-side, respectively.

Methods

Coding of EMG data

For this study, we analyzed existing data on jaw-muscle activity patterns during chewing for 32 species across 11 orders of mammals (Table 2). Some of these data consist of only raw, unquantified EMGs. In order to test the hypotheses in Table 1 with the most taxonomically and phylogenetically inclusive data set, we coded muscle timing and relative magnitudes of muscles as binary traits (i.e., 0, 1). We coded the presence or absence of Triplet I or II (Hypotheses 1, 2A–C, 3A–C, 4A and B); the ratio of relative WS to BS superficial masseter (Hypothesis 3A), medial pterygoid (Hypothesis 3B), posterior temporalis (Hypothesis 3C), and deep masseter activity (Hypothesis 5); the asynchrony or synchrony of WS and BS deep masseter (Hypotheses 4A and B); and the timing of BS deep masseter and BS posterior temporalis activity as preceding or following the WS superficial masseter (Hypotheses 5 and 6, respectively) (Table 2). A Triplet was considered to be present if the muscles comprising that Triplet (e.g., BS superficial masseter, BS medial pterygoid and WS temporalis for Triplet I) reached peak activity simultaneously or sequentially to the exclusion of peak activity in all other muscles. The ratio of recruitment from WS and BS jaw muscles (hereafter, the W/B ratio) was based on published estimates of these ratios (see references in Table 1). Here, we scored W/B ratios as either <1.5 or >1.5 based on the distribution of variation observed within primates for W/B ratios among the jaw muscles, which comprise much of the present data set. WS and BS muscle pairs were scored as either synchronous or asynchronous following Weijs’ (1994) model for mammalian jaw-muscle function. In this model, closing of the jaw from maximum gape in primitive mammals is initiated by synchronous activity of the WS and BS anterior temporalis and deep masseters, and asynchronous activity is hypothesized to be characteristic of species that have more pronounced jaw movements during fast closing and the power stroke. Muscles were considered to be synchronous if the peak activity of the WS and BS muscles were temporally closest to each other to the exclusion of peak activity of other muscles. Additionally, the WS and BS muscles on average had to reach peak activity within 5 milliseconds of each other. It is important to note, however, that we are making assumptions about muscle homology across mammals. Future work could reveal interesting patterns related to muscle vectors that do not necessarily correspond to the conventions for naming muscles and/or to biological homology (see Druzinsky et al. 2011).

Table 2.

Coding of EMG data for analyses of correlated evolution

Triplets
 W/B ratios
 WS and BS synchrony Timing relative to WSM
Triplet I Triplet II Sup Mass Med Ptery Post Temp Deep Mass Deep Mass BS Deep Mass BS Post Temp
0: absent 0: <1.5 0: synchronous 0: precedes WSM
Species 1: present 1: >1.5 1: asynchronous 1: follows WSM References
Pedetes capensis 0 0 – – – – 0 0 0 Offermans and De Vree 1993
Rattus norvegicus 0 0 – – – – 0 0 0 Weijs and Dantuma 1975
Mesocricetus auratus 1 1 – – – – 1 0 – Gorniak 1977
Cavia porcellus 0 0 – – – – 0 – 0 Byrd 1981
Aplodontia rufa 0 0 – – – – 0 – – Druzinsky 1985
Oryctolagus cuniculus 1 1 0 1 0 0 1 0 0 Weijs et al. 1989; Langenbach et al. 2001
Papio hamadryas 1 1 1 0 0 0 1 1 1 Hylander and Johnson 1994; Hylander et al. 2000; Hylander et al. 2005
Macaca fascicularis 1 1 0 0 0 0 1 1 1 Hylander and Johnson 1994; Hylander et al. 2000; Hylander et al. 2005
Homo sapiens 1 1 1 1 0 0 1 1 1 C. J. Vinyard, unpublished data
Aotus trivirgatus 0 0 0 – 0 0 0 1 1 Hylander et al. 2004; Hylander et al. 2005
Callithrix jacchus 0 0 0 0 0 0 0 1 1 Vinyard et al. 2001; C. J. Vinyard, unpublished data
Cebus apella 1 1 0 0 0 0 1 1 1 C. J. Vinyard, unpublished data
Propithecus verreauxi 1 1 0 0 0 0 1 1 1 Hylander et al. in press
Hapalemur griseus 1 1 0 – 0 1 1 1 1 W. L. Hylander, unpublished data
Lemur catta 1 – 1 – 1 1 1 0 1 Vinyard et al. 2006; C. J. Vinyard, unpublished data
Otolemur crassicaudatus 1 – 1 – 1 1 1 0 1 Hylander and Johnson 1994; Hylander et al. 2004; Hylander et al. 2005
Tupaia belangeri 1 – 1 – 1 1 1 0 1 Vinyard et al. 2005
Sus scrofa 1 1 1 – – 1 1 1 – Herring 1976; Herring 1977; Herring et al. 1979
Capra hircus 0 0 1 0 0 1 1 0 0 Williams 2004; Williams et al. 2007
Lama pacos 1 1 0 0 0 1 1 1 1 Williams 2004; Williams et al. 2007; Williams et al. 2010
Equus caballus 1 0 1 0 0 0 1 1 1 Williams 2004; Williams et al. 2007
Canis lupus – – 0 – 1 – – – – Dessem 1989
Felis silvestris – – – – – – 0 – – Gorniak and Gans 1980
Pteropus giganteus 0 0 0 0 – – 0 1 – De Gueldre and De Vree 1988
Myotis lucifugus 0 0 – – – – 1 0 – Kallen and Gans 1972
Tenrec ecaudatus – – – – – – 1 – – Oron and Crompton 1985
Macropus eugenii 0 0 0 0 0 – 1 – 0 Crompton et al. 2008
Macropus rufus 0 0 0 0 0 – 1 – 0 Crompton et al. 2008
Potorous tridactylus 1 1 1 0 0 0 1 1 0 A. W. Crompton, unpublished data
Lasiorhinus latifrons 0 0 1 1 1 1 1 – – Crompton et al. 2008
Phascolarctos cinereus 0 0 1 1 1 1 1 1 0 Crompton et al. 2010
Didelphis virginiana 1 1 0 – – 1 1 0 – Crompton and Hylander 1986
Open in a new tab

Analysis of correlated evolution

To analyze the correlated evolution between pairs of traits, we first “pruned” the phylogenetic tree published by Bininda-Emonds et al. (2007) using Mesquite Ver. 2.73 (Maddison and Maddison 2010). The tree was pruned to contain the 32 species for which we have corresponding EMG data (Fig. 1). We combined EMG data from Macaca fascicularis and Macaca fuscata and used the phylogenetic position of M. fascicularis in the phylogeny. We also used the phylogenetic position of Papio hamadryas for P. anubis, the only Papio species for which there are EMG data.

Fig. 1.

Fig. 1

Open in a new tab

Phylogeny used in the present analysis, based on data in Bininda-Emonds et al. (2007).

We then used the BayesDiscrete module within BayesTraits to fit two continuous-time Markov models to the discrete trait data in each hypothesis. In one model, traits evolve independently on the phylogeny. In the second model, traits evolve dependently and thus, are considered correlated (see Pagel 1994; Pagel and Meade 2006). In the dependent model, there can be four states, each of which is a potential combination of the two binary traits (i.e., 0,0; 1,0; 0,1; 1,1). The fit of the dependent and independent models were analyzed in two ways. First, we used a maximum likelihood method to determine whether two motor-pattern traits are correlated by calculating the likelihood ratio statistic (LR). The LR statistic is calculated as twice the difference in log-likelihoods of the dependent and independent models {i.e., 2[log-likelihood (dependent model)–log-likelihood (independent model)]}. The LR approximates a χ2 distribution with four degrees of freedom. When the log-likelihood of the dependent model is significantly greater than the log-likelihood of the independent model, there is support for correlated evolution of the two traits.

The second analysis of the fit of the dependent and independent models was conducted in BayesTraits using Bayesian Markov Chain Monte Carlo (MCMC) statistics to account for uncertainty in the phylogeny (see Pagel and Meade 2006). Bayesian MCMC statistics calculate the total harmonic mean of the maximum likelihoods of the two models. The harmonic mean of the maximum likelihood is an approximation of the marginal likelihood (as opposed to the maximum likelihood) of a model. Because we had no a priori reason to predict that a particular value is more likely, we used uninformative, uniform priors in each model. As recommended, we allowed a burn-in of 50,000 iterations and sampled every 100th step from a total of 5,050,000 iterations for dependent and independent models. From these results, we calculated twice the difference in the harmonic means of log-likelihoods between models. This difference yields a Bayes Factor, which was interpreted as support of a model of correlated evolution on the following scale: >2 = positive evidence for correlated evolution; >5 = strong evidence for correlated evolution; >10 = very strong evidence for correlated evolution (Pagel and Meade 2006; www.evolution.rdg.ac.uk).

Results

The maximum likelihood and MCMC analyses generally yield consistent results for each of the hypotheses tested and are discussed collectively (Table 3). There is very strong support for the co-evolution of Triplets I and II (Hypothesis 1). Both analyses provided their highest likelihood ratio (35.33, P < 0.001) and Bayes Factor (21.42), respectively for this test of the hypothesis.

Table 3.

Results of maximum and MCMC analyses of correlated evolution

Maximum likelihood
 MCMC
Likelihood
 Harmonic mean of log likelihood
Hypothesis Variables Dependent model Independent model Likelihood ratio P-value1 Dependent model Independent model Bayes factor2
1 Triplet I/Triplet II −18.04 −35.70 35.33 0.001 −27.18 −37.89 21.42
2A Triplet I/SM W/B Ratio −33.05 −34.34 2.58 0.630 −36.20 −36.45 0.49
2B Triplet I/MP W/B Ratio −26.14 −26.60 0.91 0.924 −29.95 −29.72 −0.46
2C Triplet I/PT W/B Ratio −29.53 −29.89 0.73 0.948 −33.03 −32.74 −0.57
3A Triplet II/SM W/B Ratio −31.96 −32.75 1.59 0.810 −34.68 −34.96 0.56
3B Triplet II/MP W/B Ratio −24.59 −25.01 0.84 0.933 −27.50 −27.43 −0.15
3C Triplet II/PT W/B Ratio −27.48 −28.30 1.64 0.801 −31.03 −30.96 −0.14
4A WS and BS DM Synchrony/Triplet I −26.90 −35.27 16.75 0.002 −33.53 −38.48 9.90
4B WS and BS DM Synchrony/Triplet II −26.92 −33.68 13.53 0.009 −32.27 −36.28 8.02
5 BS DM Timing/DM W/B Ratio −22.44 −25.92 6.95 0.139 −29.12 −30.51 2.78
6 BS DM Timing/BS PT Timing −23.28 −26.37 6.19 0.186 −29.44 −31.33 3.79
Open in a new tab

1Significance is based on a χ2 distribution with 4 df (see Pagel 1994).

2Positive Bayes Factor values support a model of dependent (correlated) evolution on the following scale: >2 = positive evidence for correlated evolution; >5 = strong evidence for correlated evolution; >10 = very strong evidence for correlated evolution (www.evolution.rdg.ac.uk).

We found no support for Hypotheses 2 and 3, focusing on the correlated evolution of W/B ratios and the presence or absence of Triplets. In several cases, the Bayes Factors was negative (Hypotheses 2B and C, 3B and C) (Table 3). In the remaining two cases (Hypotheses 2A and 3A), the Bayes Factor was positive, but well below the cutoff of two, thereby indicating no support. As all of the LRs were not significant as well, these results unequivocally refute a model of correlated evolution between Triplets and relative levels of muscle activity.

We observed support for the correlated evolution of BS deep masseter timing and the loss of both Triplets (Hypotheses 4A and B). In both tests, the LR ratio was significant (P < 0.01) and the Bayes Factor >5 indicating strong evidence for correlated evolution of WS and BS deep masseter asynchrony and Triplets I and II.

We found mixed support for Hypotheses 5 and 6, relating the evolution of timing in the BS deep masseter to both deep masseter W/B ratios and timing in the BS posterior temporalis (Table 3). In both cases, the LR was not significant but the MCMC analysis did suggest a trend toward co-evolution of the two traits with Bayes Factors >2.

Discussion

The evolution of Triplets, BS deep masseter activity, and mammalian mastication

The number and complex anatomical organization of the mammalian jaw-closing muscles create the potential for numerous combinations of recruitment patterns to produce movements and bite forces during mastication. However, several researchers note that in most mammals studied to date, jaw closing is primarily driven by activity in Triplet I muscles followed by the activity in Triplet II muscles (e.g., Herring 1976; Weijs 1994). Given that Triplet I and Triplet II muscles are “mirror images” of each other, it is not surprising that our results lend support to Hypothesis 1 that the Triplets co-evolved. If one group of muscles becomes active first to pull the jaw towards the working side during fast-closing, then the remaining muscles are left to pull the jaw towards the balancing side through the power stroke. This means that Triplets I and II are both present in many clades, most notably Tupaia, strepsirrhine and most anthropoid primates, some marsupials and all artiodactyls (Fig. 2). Didelphis is often used as a model for the primitive mammalian chewing motor pattern because of its primitive masticatory morphology (Crompton and Hylander 1986), and therefore, the presence of the Triplets in this species suggests that they were likely present in primitive mammals.

Fig. 2.

Fig. 2

Open in a new tab

Mirrored phylogeny from Fig. 1 showing the distribution of Triplet I and II.

While retaining the Triplets, Phascolarctos, the koala, has evolved a chewing motor pattern that most closely resembles that of alpacas among herbivorous placentals, which shows a modification from the primitive Triplet pattern for transverse chewing (Williams et al. 2007; Crompton et al. 2010). Like alpacas, koalas have a derived molar morphology specialized for transverse shearing of vegetation (Crompton et al. 2010). Among rodents, it is noteworthy that Triplets I and II re-evolved together in a clade that, according to Weijs (1994), has diverged significantly from the Triplet pattern. Specifically, two non-Triplet-like motor patterns consisting of either bilaterally synchronous activity of the jaw adductors (“rodent symmetric” motor pattern) or alternating activity on the left and right sides (“rodent alternate” motor pattern) have been observed in other rodents. Thus, reappearance of the Triplet motor pattern in Mesocricetus, is interesting, and it may be that this pattern is more pervasive among rodents than previously thought.

Although many primates and marsupials have retained the Triplets from a primitive condition, there are several examples in which chewing motor patterns have diverged from this state (Fig. 2). For example, in the New World monkeys Callithrix jacchus and Aotus trivirgatus, the BS superficial masseter of Triplet I fires with the WS superficial masseter of Triplet II (Vinyard et al. 2001, C. J. Vinyard, unpublished data; Hylander et al. 2004, 2005). Altogether different groupings of jaw adductors have been observed in macropodines (Macropus spp.) and wombats (Lasiorhinus latifrons) among marsupials (Crompton et al. 2008a, 2008b). Wombats arguably have the most unique pattern of coordination of jaw muscles because jaw movements during closing are entirely driven by the WS muscles, with little to no BS activity (Crompton et al. 2008b). Crompton et al. (2008b) demonstrated that this motor pattern is related to the derived morphology of the masticatory apparatus of wombats, particularly in the mandible as well as in the attachments and orientations of the jaw adductors. A final exception to the Triplet pattern is the horse (Equus caballus). In horses, the BS medial pterygoid fires with the WS superficial masseter and WS medial pterygoid of Triplet II rather than with Triplet I muscles. Thus, this species does not appear to retain the Triplets, although the reason for this shift in activity is not clear (Williams et al. 2007).

We did not observe the correlated evolution of Triplet I and II with either an increase or decrease in W/B ratios (Hypotheses 2 and 3). These analyses were primarily exploratory, but we initially speculated that high W/B ratios (i.e., >1.5) for the superficial masseter and medial pterygoid and low W/B ratios for the posterior temporalis would correlate with the presence of both Triplets. Our predictions follow from Triplet I muscles being active early during fast-closing through the initiation of the power stroke whereas Triplet II muscles exhibit pronounced activity during the majority of the power stroke to break down food. As Triplet II consists of the WS superficial masseter, WS medial pterygoid and BS posterior temporalis, these muscles may need to be recruited more heavily during the power stroke compared to the contralateral muscles during fast-closing (i.e., higher superficial masseter and medial pterygoid versus low posterior temporalis W/B ratios). However, even in the most data-rich clades, the relationship between the presence/absence of Triplets I or II and the W/B ratio is not consistent. For example, all strepsirrhine primates exhibit Triplet I but only Lemur and Otolemur have W/B ratios for the superficial masseter of >1.5 (Fig. 3). Triplet I muscles in these species may be exhibiting pronounced recruitment, resulting in high superficial masseter and medial pterygoid W/B ratios (i.e., >1.5), simply as a means to increase occlusal forces during the chewing of hard or tough foods. The remaining two strepsirrhine species (Hapalemur and Propithecus) have superficial masseter W/B ratios <1.5. In marsupials, Didelphis, Phascolarctos, and Potorous have Triplet I but only the latter two species exhibit superficial masseter W/B ratios >1.5. Lasiorhinus also has a superficial masseter W/B ratio >1.5 but lacks Triplet I. These results highlight the numerous combinations of muscle recruitment patterns that facilitate mastication and suggest that the evolution of W/B ratios occurred within the consistent presence of the Triplets across many mammalian clades.

Fig. 3.

Fig. 3

Open in a new tab

Mirrored phylogeny from Fig. 1 showing the distribution of Triplet I and superficial masseter W/B ratios.

Hypothesis 4 focuses on the evolution of the Triplets and synchrony in the WS and BS deep masseter muscles. Weijs (1994) proposed that mammals may enhance the transverse component to jaw movements during fast-closing and the power stroke by temporally separating the activity of the WS and BS deep masseters. The WS deep masseter muscle assists Triplet I in drawing the WS dentary laterally during fast-closing whereas the BS deep masseter assists in drawing the WS dentary back towards the midline during the power stroke. Here, we found overwhelming support for the correlated evolution of deep masseter WS and BS asynchrony and the evolution of Triplets I and II. However, these results may be due to the fact that our data set consists almost entirely of mammals that have a transverse component to the power stroke. For example, asynchrony in the WS and BS deep masseters is stereotypical for ungulates and primates despite variation in other aspects of their motor patterns (Hylander and Johnson, 1994; Hylander et al. 2000; Williams et al. 2007). Additional data from carnivores with limited transverse movements and from rodents exhibiting propalinal mastication would be useful for clarifying the strength of this relationship across mammals. We do not, however, expect a reversal in the present trend, based on our existing understanding of the function of the jaw muscles in these groups.

Finally, we also found positive evidence for the correlated evolution between a late-acting BS deep masseter (i.e., following peak activity of the WS superficial masseter) and a decrease in W/B ratio to <1.5 (Hypothesis 5; Fig. 4) as well as a late-acting BS posterior temporalis (i.e., following the WS superficial masseter) (Hypothesis 6; Fig. 5). Co-evolution of a delay in activity of the BS deep masseter and a decrease in the W/B ratio of the deep masseter is hypothesized to relate to ossification, or fusion of the mandibular symphysis, as well as to other changes in jaw morphology in primates (see Hylander et al. 2000, 2004 and references therein; Vinyard et al. 2011). Notable exceptions to this hypothesis occur in primates (e.g., Hapalemur griseus) (W. L. Hylander, unpublished data) and artiodactyls (e.g., Lama pacos) (Williams et al. 2007, 2010). Both have a BS deep masseter that peaks after the WS superficial masseter, but a deep masseter W/B ratio that is >1.5. These species also differ in symphyseal fusion—Hapalemur has an unfused symphysis whereas Lama has a fused symphysis. The functional significance of this link between motor pattern and morphology is further explored by Vinyard et al. (2011). Correlated evolution between the timing of the BS deep masseter and BS posterior temporalis also received moderate support. It is not readily apparent why these two traits co-evolved although we hypothesize that delaying both muscles increases transverse movement and force at the end of the power stroke (see Hylander et al. 2005; Williams et al. 2007). In horses, it has been hypothesized that this combination of muscles may help overcome the inertia of their massive jaws at the end of the power stroke when the other jaw adductors are decreasing their activity (Williams et al. 2007). Future work aimed at linking recruitment patterns to jaw movements across mammals can test this hypothesis.

Fig. 4.

Fig. 4

Open in a new tab

Mirrored phylogeny from Fig. 1 showing the distribution of BS deep masseter timing and deep masseter W/B ratios.

Fig. 5.

Fig. 5

Open in a new tab

Mirrored phylogeny from Fig. 1 showing the distribution of BS deep masseter and BS posterior temporalis timing relative to the WS superficial masseter.

Caveats and future research

Although some of the results provide strong support for the correlated evolution of several features of the mammalian chewing motor pattern, the present study has several caveats. First, the EMG data were collected and analyzed using different techniques. However, by treating the data as binary states, we hope that we have reduced variation associated with combining these data sets from different laboratories. Our criterion for accepting data into the study was based on our interpretation of published data for which the methods of collecting data were well described. When we did not have direct access to the data, we also utilized the interpretations of the data in publications. Thus, some of the coding of states was subjective and relied on descriptive analyses of the motor patterns by the primary investigators (e.g., Oron and Crompton 1985; De Gueldre and De Vree 1988). We were conservative, however, and chose not to include many of the existing data because we could not adequately assess the variable states. Future efforts to digitize and re-analyze existing data for accession into the FEED database will allow us to more accurately code some of the species and to fill out the data set where we have missing data. Fortunately, a rather large subset of the EMG data was collected and quantified using comparable methods. These include all data from primates, goats, alpacas, horses and treeshrews. Moreover, the studies of marsupials utilized a method for calculating timing differences between muscles similar to that used of the present study.

By far the greatest limitation of the present study is the limited taxonomic, morphologic, and phylogenetic diversity in our sample. This is not an insurmountable limitation but it requires a multi-pronged approach. We first need to digitize existing EMG data from several different laboratories. Some of these data are important as they would allow us to code states currently missing for some species in the present study as well as increase the representation of mammalian orders that we could not include here (e.g., Hyracoidea). Additionally, and perhaps more importantly, we need better taxonomic sampling of feeding physiology across the mammalian phylogeny. The existing physiological data and the present study are heavily weighted towards ungulates, primates and marsupials. While these data still only represent a small portion of the total number of species in each of these orders, they do span some taxonomic and morphological breadth within each clade. In contrast, some of the most morphologically diverse and/or speciose clades are represented by only a few species, such as rodents, bats and carnivores. Additionally, we have no data from some of the more derived clades of mammals, and this may prove to be very important for confirming or refuting the patterns observed here. For example, loss of complex occlusion and/or teeth in some species (e.g., termite and ant specialists, some marine mammals) is derived for mammals, yet the Triplets may be primitive. It would be interesting to know whether some of these species retain the Triplets with reduced asynchrony resulting in reduced transverse movement during jaw closing, as proposed by the present study for carnivorous species. Alternatively, increased sampling, particularly focusing on closely related species with markedly divergent diets and/or masticatory morphology, may reveal new motor patterns, as has been observed recently in marsupials (Crompton et al. 2008a, 2008b; 2010).

Finally, we need a better understanding of the evolutionary relationships between chewing motor patterns, the kinematics of the chewing cycle, and the morphology of the masticatory apparatus. The extent to which these three are correlated has not been rigorously tested in a phylogenetically broad context. As discussed above, some of the changes in chewing motor patterns across mammals are associated with changes in morphology of the jaw-muscles, jaws and teeth. A major role of the jaw adductors is to produce movement during chewing, but quantitative data on kinematics of the jaw are scarce. Thus, we do not know the extent to which subtle differences in motor patterns of the jaw muscles result in quantifiable kinematic difference in jaw movements between species. Here, however, these subtle differences may be represented as a state change in the motor pattern (e.g., shift in the BS deep masseter relative to the WS superficial masseter). At present, we have a better link between chewing motor patterns and morphology in mammals, as this has been the subject of several previous studies (e.g., Hylander et al. 2000; Williams et al. 2008; Crompton et al. 2010). The article by Vinyard et al. (2011) is a first step towards broader phylogenetic analyses of chewing motor patterns and morphology, and like the present study, supports the idea that chewing motor patterns are one component of an evolving masticatory system throughout mammalian evolution.

Funding

Funds for the symposium were generously provided by National Science Foundation (IOS-1050313 to S.H.W., C.J.V., C.E.W., and R.Z.G.); the Society for Comparative and Experimental Biology Divisions of Vertebrate Morphology, Comparative Physiology and Biochemistry, Neurobiology, and Systematic and Evolutionary Biology. Research presented in this article was supported by the National Science Foundation (IOS-0520855 to S.H.W., BCS-0552285 and BCS-0412153 to C.J.V., BCS-0094522 to C.E.W., and BCS-0138565, SBR-9429764, BNS-9100523 and BCS-0241652 to W.L.H.); the National Institutes of Health (DE04531, DE05595 and DE05663); the Australian Research Council (to A.W.C.).

Acknowledgments

This article was presented in the 2011 SICB symposium “Synthesis of Physiologic Data from the Mammalian Feeding Apparatus using FEED: The Feeding Experiments End-User Database.”

References

  1. Bininda-Emonds OR, Cardillo M, Jones KE, MacPhee RD, Beck RM, Grenyer R, Price SA, Vos RA, Gittleman JL, Purvis A. The delayed rise of present-day mammals. Nature. 2007;446:507–12. doi: 10.1038/nature05634. [DOI] [PubMed] [Google Scholar]
  2. Byrd KE. Mandibular movement and muscle activity during mastication in the guinea pig (Cavia porcellus) J Morphol. 1981;170:147–69. doi: 10.1002/jmor.1051700203. [DOI] [PubMed] [Google Scholar]
  3. Crompton AW. Masticatory motor programs in Australian herbivorous mammals: diprotodontia. Integr Comp Biol. 2011;51:271–81. doi: 10.1093/icb/icr028. [DOI] [PubMed] [Google Scholar]
  4. Crompton AW, Barnet J, Lieberman DE, Owerkowicz T, Skinner J, Baudinette RV. Control of jaw movements in two species of macropodines (Macropus eugenii and Macropus rufus) Comp Biochem Physiol Part A. 2008a;150:109–23. doi: 10.1016/j.cbpa.2007.10.015. [DOI] [PubMed] [Google Scholar]
  5. Crompton AW, Hylander WL. Changes in mandibular function following the acquisition of a dentary-squamosal jaw articulation. In: Hotton M, MacLean P, Roth J, Roth E, editors. Ecology and biology of mammal-like reptiles. Washington, DC: Smithsonian Institution Press; 1986. pp. 263–87. [Google Scholar]
  6. Crompton AW, Lieberman DE, Owerkowicz T, Baudinette RV, Skinner J. Motor control of masticatory movements in the Southern hairy-nosed wombat (Lasiorihinus latifron) In: Vinyard CJ, Ravosa MJ, Wall CE, editors. Primate craniofacial function and biology. New York: Springer; 2008b. pp. 83–111. [Google Scholar]
  7. Crompton AW, Owerkowicz T, Skinner J. Masticatory motor pattern in the koala (Phascolarctos cinereus): a comparison of jaw movements in marsupial and placental herbivores. J Exp Zool A Ecol Genet Physiol. 2010;313:564–78. doi: 10.1002/jez.628. [DOI] [PubMed] [Google Scholar]
  8. De Gueldre G, De Vree F. Quantitative electromyography of the masticatory muscles of Pteropus giganteus (Megachiroptera) J Morphol. 1988;196:73–106. doi: 10.1002/jmor.1051960107. [DOI] [PubMed] [Google Scholar]
  9. Deban SM, O’Reilly JC, Nishikawa KC. The evolution of the motor control of feeding in amphibians. Am Zool. 2001;41:1280–98. [Google Scholar]
  10. Dessem D. Interactions between jaw-muscle recruitment and jaw-joint forces in Canis familiaris. J Anat. 1989;164:101–21. [PMC free article] [PubMed] [Google Scholar]
  11. Druzinsky RE. Anatomy and EMG of the masseter of Aplodontia rufa. Fortschr Zool. 1985;30:281–3. [Google Scholar]
  12. Druzinsky R, Doherty A, De Vree F. Mammalian masticatory muscles: homology, nomenclature, and diversification. Integr Comp Biol. 2011;51:224–34. doi: 10.1093/icb/icr067. [DOI] [PubMed] [Google Scholar]
  13. Gorniak GC. Feeding in golden hamsters, Mesocricetus auratus. J Morphol. 1977;154:427–58. doi: 10.1002/jmor.1051540305. [DOI] [PubMed] [Google Scholar]
  14. Gorniak GC, Gans C. Quantitative assay of electromyograms during mastication in domestic cats (Felis catus) J Morphol. 1980;163:253–81. doi: 10.1002/jmor.1051630304. [DOI] [PubMed] [Google Scholar]
  15. Herring SW. The dynamics of mastication in pigs. Arch Oral Biol. 1976;21:473–80. doi: 10.1016/0003-9969(76)90105-9. [DOI] [PubMed] [Google Scholar]
  16. Herring SW. Mastication and maturity: a longitudinal study in pigs. J Dent Res. 1977;56:1377–82. doi: 10.1177/00220345770560111701. [DOI] [PubMed] [Google Scholar]
  17. Herring SW. The ontogeny of mammalian mastication. Am Zool. 1985;25:339–49. [Google Scholar]
  18. Herring SW. Muscles of Mastication: architecture and functional organization. In: Davidovitch Z, editor. The biological mechanisms of tooth movement and craniofacial adaptation. Columbus: The Ohio State University College of Dentistry; 1992. pp. 541–8. [Google Scholar]
  19. Herring SW. Functional morphology of mammalian mastication. Am Zool. 1993;33:289–99. [Google Scholar]
  20. Herring SW, Grimm AF, Grimm BR. Functional heterogeneity in a multipinnate muscle. Am J Anat. 1979;154:563–76. doi: 10.1002/aja.1001540410. [DOI] [PubMed] [Google Scholar]
  21. Hiiemae K. Feeding in mammals. In: Schwenk K, editor. Feeding: form, function and evolution in tetrapod vertebrates. New York: Academic Press; 2000. pp. 411–48. [Google Scholar]
  22. Hiiemae KM. Masticatory movements in primitive mammals. In: Anderson DJ, Matthews B, editors. Mastication. Bristol: John Wright and Sons, Ltd; 1976. pp. 105–18. [Google Scholar]
  23. Hiiemae KM. Mammalian mastication: a review of the activity of the jaw muscles and the movements they produce in chewing. In: Butler P, Joysey K, editors. Studies on the development, structure and function of teeth. London: Academic Press; 1978. pp. 359–98. [Google Scholar]
  24. Hylander WL, Johnson KR. Jaw muscle function and wishboning of the mandible during mastication in macaques and baboons. Am J Phys Anthropol. 1994;94:523–47. doi: 10.1002/ajpa.1330940407. [DOI] [PubMed] [Google Scholar]
  25. Hylander WL, Ravosa MJ, Ross CF, Wall CE, Johnson KR. Symphyseal fusion and jaw-adductor muscle force: an EMG study. Am J Phys Anthropol. 2000;112:469–92. doi: 10.1002/1096-8644(200008)112:4<469::AID-AJPA5>3.0.CO;2-V. [DOI] [PubMed] [Google Scholar]
  26. Hylander WL, Vinyard CJ, Ravosa MJ, Ross CF, Wall CE, Johnson KR. Jaw adductor force and symphyseal fusion. In: Anapol F, German RZ, Jablonski NG, editors. Shaping primate evolution: papers in honor of Charles Oxnard. Cambridge: Cambridge University Press; 2004. pp. 229–57. [Google Scholar]
  27. Hylander WL, Vinyard CJ, Wall CE, Johnson KR. Functional and evolutionary significance of the recruitment and firing patterns of the jaw adductors during chewing in Verreauxi’s sifaka (Propithecus verreauxi) Am J Phys Anthropol. in press doi: 10.1002/ajpa.21529. [DOI] [PubMed] [Google Scholar]
  28. Hylander WL, Wall CE, Vinyard CJ, Ross C, Ravosa MR, Williams SH, Johnson KR. Temporalis function in anthropoids and strepsirrhines: an EMG study. Am J Phys Anthropol. 2005;128:35–56. doi: 10.1002/ajpa.20058. [DOI] [PubMed] [Google Scholar]
  29. Kallen FC, Gans C. Mastication in the little brown bat, Myotis lucifugus. J Morphol. 1972;136:385–420. doi: 10.1002/jmor.1051360402. [DOI] [PubMed] [Google Scholar]
  30. Langenbach GE, Weijs WA, Brugman P, van Eijden TM. A longitudinal electromyographic study of the postnatal maturation of mastication in the rabbit. Arch Oral Biol. 2001;46:811–20. doi: 10.1016/s0003-9969(01)00043-7. [DOI] [PubMed] [Google Scholar]
  31. Maddison W, Maddison D. Mesquite: a modular system for evolutionary analysis. 2010 Version 2.73 ( http://mesquiteproject.org) [Google Scholar]
  32. Offermans M, De Vree F. Electromyography and mechanics of mastication in the springhare, Pedetes Capensis (Rodentia, Pedeitdae) Belg J Zool. 1993;123:231–60. [Google Scholar]
  33. Oron U, Crompton AW. A cineradiographic and electromyographic study of mastication in Tenrec ecaudatus. J Morphol. 1985;185:155–82. doi: 10.1002/jmor.1051850203. [DOI] [PubMed] [Google Scholar]
  34. Pagel M. Detecting correlated evolution on phylogenies – a general-method for the comparative-analysis of discrete characters. Proc Roy Soc Lond B. 1994;255:37–45. [Google Scholar]
  35. Pagel M, Meade A. Bayesian analysis of correlated evolution of discrete characters by reversible-jump Markov chain Monte Carlo. Am Nat. 2006;167:808–25. doi: 10.1086/503444. [DOI] [PubMed] [Google Scholar]
  36. Smith KK. Are neuromotor systems conserved in evolution? Brain, Beh Evol. 1994;43:293–305. doi: 10.1159/000113641. [DOI] [PubMed] [Google Scholar]
  37. Vinyard CJ, Wall CE, Williams SH, Johnson KR, Hylander WL. Masseter electromyography during chewing in ring-tailed lemurs (Lemur catta) Am J Phys Anthropol. 2006;130:85–95. doi: 10.1002/ajpa.20307. [DOI] [PubMed] [Google Scholar]
  38. Vinyard CJ, Williams SH, Wall CE, Doherty AH, Cromptom AW, Hylander WL. A preliminary analysis of correlations between chewing motor patterns and mandibular morphology across mammals. Integr Comp Biol. 2011;51:260–70. doi: 10.1093/icb/icr066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Vinyard CJ, Williams SH, Wall CE, Johnson KR, Hylander WL. Deep masseter recruitment patterns during chewing in callitrichids. Am J Phys Anthropol Supplement. 2001;32:156. [Google Scholar]
  40. Vinyard CJ, Williams SH, Wall CE, Johnson KR, Hylander WL. Jaw-muscle electromyography during chewing in Belanger’s Treeshrews (Tupaia belangeri) Am J Phys Anthropol. 2005;127:26–45. doi: 10.1002/ajpa.20176. [DOI] [PubMed] [Google Scholar]
  41. Wainwright PC, Friel JP. Behavioral characters and historical properties of motor patterns. In: Wagner GP, editor. The character concept in evolutionary biology. New Haven: Academic Press; 2001. pp. 285–301. [Google Scholar]
  42. Weijs WA. Evolutionary approach of masticatory motor patterns in mammals. In: Bels V, Chardon M, Vandewalle P, editors. Adv Comp Environ Physiol. Berlin: Springer-Verlag; 1994. pp. 282–320. [Google Scholar]
  43. Weijs WA, Brugman P, Grimbergen CA. Jaw movements and muscle activity during mastication in growing rabbits. Anat Rec. 1989;224:407–16. doi: 10.1002/ar.1092240309. [DOI] [PubMed] [Google Scholar]
  44. Weijs WA, Dantuma R. Electromyography and mechanics of mastication in the albino rat. J Morphol. 1975;146:1–33. doi: 10.1002/jmor.1051460102. [DOI] [PubMed] [Google Scholar]
  45. Williams SH. Mastication in selenodont artiodactyls: an in vivo study of masticatory form and function in goats and alpacas [PhD] Durham: Duke University; 2004. p. 302. [Google Scholar]
  46. Williams SH, Sidote JV, Stover KK. Occlusal development and masseter activity in alpacas (Lama pacos) Anat Rec. 2010;293:126–34. doi: 10.1002/ar.21016. [DOI] [PubMed] [Google Scholar]
  47. Williams SH, Vinyard CJ, Wall CE, Hylander WL. Masticatory motor patterns in ungulates: a quantitative assessment of jaw-muscle coordination in goats, alpacas and horses. J Exp Zool Part A Ecol Genet Physiol. 2007;307:226–40. doi: 10.1002/jez.362. [DOI] [PubMed] [Google Scholar]
  48. Williams SH, Wall CE, Vinyard CJ, Hylander WL. Symphyseal fusion in selenodont artiodactyls: new insights from in vivo and comparative data. In: Vinyard CJ, Ravosa MR, Wall CE, editors. Primate craniofacial function and biology. New York: Springer; 2008. pp. 39–61. [Google Scholar]

Articles from Integrative and Comparative Biology are provided here courtesy of Oxford University Press

RESOURCES