Diet drove brain and dental morphological coevolution in strepsirrhine primates.

The evolution of the remarkably complex primate brain has been a topic of great interest for decades. Multiple factors have been proposed to explain the comparatively larger primate brain (relative to body mass), with recent studies indicating diet has the greatest explanatory power. Dietary specialisations also correlate with dental adaptations, providing a potential evolutionary link between brain and dental morphological evolution. However, unambiguous evidence of association between brain and dental phenotypes in primates remains elusive. Here we investigate the effect of diet on variation in primate brain and dental morphology and test whether the two anatomical systems coevolved. We focused on the primate suborder Strepsirrhini, a living primate group that occupies a very wide range of dietary niches. By making use of both geometric morphometrics and dental topographic analysis, we extend the study of brain-dental ecomorphological evolution beyond measures of size. After controlling for allometry and evolutionary relatedness, differences in brain and dental morphology were found between dietary groups, and brain and dental morphologies were found to covary. Historical trajectories of morphological diversification revealed a strong integration in the rates of brain and dental evolution and similarities in their modes of evolution. Combined, our results reveal an interplay between brain and dental ecomorphological adaptations throughout strepsirrhine evolution that can be linked to diet.

Introduction

Larger-than-expected brain size (relative to body mass and compared to other mammals) is commonly interpreted as evidence for increased intelligence as an evolutionary novelty in primates [1-3]. Understanding the factors that shaped the evolution of the relatively large primate brain has been a topic of extensive debate, with multiple competing hypotheses postulated over the years [4]. Diet, increased social complexity and the energetic cost of brain tissue development and maintenance have been studied as predictors of brain size in primates [5-7]. Accumulating evidence points to diet as a main predictor of brain size across Primates [4, 8, 9], with factors like sociality playing an important role in certain groups (i.e. anthropoids, [10]). Increasing degrees of frugivory in particular has been identified as a major factor linked to brain enlargement in primates [4]. Frugivory has been interpreted both as a selective pressure (increased spatial information storage and cognitive demands of “extractive foraging”; [11]) and as an unconstraint (enabling higher energy allocation and turnover; [6]) on brain tissue enlargement.

Studies have discussed the limitations of interpreting brain ecomorphological adaptations solely based on size [12], reflecting the compartmentalised specialisation of the brain and highlighting the need to study brain morphology in a way that also takes shape information into consideration [3, 13, 14]. Understanding how functionally distinct brain structures change differentially in association with ecological (e.g. diet or social complexity) and biological (e.g. allometry) traits can provide a more comprehensive perspective on the eco-evolutionary dynamics of brain evolution [12-15]. Though ecology-driven variations in functionally distinct brain regions and overall brain size may face ontogenetic and functional constraints [16], there is considerable evidence that changes in the morphology of specific brain regions reflect functional adaptions [14, 17] ultimately influencing brain morphology. Parallel lineage-specific and ecology-based patterns of morphological adaptation across brain regions reflect the mosaic evolution of the primate brain as being linked to socioecological differences [14]. Moreover, independent variation patterns between brain size and other phenotypic traits suggest primate brain evolution followed a multi-patterned trajectory [13, 15]. Geometric morphometrics has proved a successful tool at capturing multiple dimensions of brain morphology to reconstruct macroevolutionary processes [13, 15].

Dental morphological variation is also associated with dietary adaptations, providing a potential link between brain and dental ecomorphological evolution [18, 19]. Studies on primate brain-dental functional coevolution have focused on hominins, hypothesising increased dietary quality is associated with a reduction in postcanine dentition and an increase in brain size, resulting in the markedly derived Homo large brain and small teeth [18-21]. Recent studies have debated the validity of the inferred coevolution of brain and dental function [18, 19]. Accounting for allometry and evolutionary relatedness reduces the signal of a diet-based brain-dental coevolution across primates, with the exception of prosimians and platyrrhines [18]. Moreover, rates of brain and dental morphological evolution in hominins seem to have followed independent trajectories, indicating decoupled evolution [19]. However, similar to the limitations that come with analysing brain size alone, limiting the study of dental morphological evolution to changes in size neglects other dimensions of dental morphology that have been shown to experience selective pressures [22-24]. Dental topographic analysis has provided novel insights into dental ecomorphological adaptations and evolution, and provides a quantitative toolkit for describing dental shape [25].

Stemming from an inferred frugivorous common ancestor, primates of the suborder Strepsirrhini underwent one of the most impressive adaptive radiations among living primates [26], coupled with significant ecomorphological diversification [25]. The primate suborder Strepsirrhini diverged from haplorrhines approximately 60Ma [26]. Comprising more than 120 living species, they exhibit intermediate relative brain size between non-primate mammals and more derived anthropoids [8, 9], and a variety of dietary and foraging specialisations that are reflected in their dental morphology [25-27]. Strepsirrhines are thought to have diversified in continental Africa, experiencing an early Oligocene partial extinction event, followed by colonisation of Madagascar, perhaps in two independent events [28, 29]. Once in Madagascar, strepsirrhines diversified to fill a range of dietary niches (including folivory, frugivory, gummivory and insectivory) and specialise for different activity periods [25]. A remarkable example of the strepsirrhine diversification is the Aye-Aye (Daubentonia madagascariensis) and its combination of morphological traits (e.g. ever-growing incisors, elongated middle digit and squirrel-like skull) associated to its unique percussive foraging strategy [30]. Studies have suggested slower evolutionary rates and a lack of allometry in strepsirrhine primate brain shape [13], and a positive correlation between postcanine teeth size and dietary quality [18]. However, given the highly derived morphology of modern catarrhines (and hominins in particular), it is possible that complex evolutionary processes in Strepsirrhini are obscured when analysing Primates as a whole.

In this study, we investigate the evolution of strepsirrhine brain and dental morphology to test whether diet acted as a link to the integration of brain and dental morphological diversification. To draw a clear link between brain and dental evolution and diet, we account for the effect of allometry and evolutionary relatedness in morphological variation. First, we quantify brain and dental morphology using geometric morphometrics and dental topographic analysis. We decompose brain morphology into its size and shape components, exploring decoupled brain ecomorphological adaptations. After accounting for allometry and phylogenetic structuring, we assess diet-based differences in brain and dental morphologies and test the integration between them. Because diet is closely related to body mass, and also tracks clade membership to some degree, we anticipate that controlling for these factors may remove any effect of diet. Second, we quantify per-dietary group and per-species rates of brain and dental evolution, test for differences between dietary groups and estimate the integration between rates of brain and dental evolution. Finally, we fit competing models of morphological evolution and compare them across dietary groups. Based on the apparently decoupled evolution of brain enlargement and dental reduction in other primate groups that have been assessed, we do not expect to find a strong correlation in evolutionary rates in strepsirrhines.

Materials and methods

Data

Data on body mass (BdM), brain mass (BrM) and endocranial volume (ECV) was gathered from the literature at species level, body and brain mass data being averages of mixed-sex samples [4, 8, 9, 31]. Species were assigned to one of three dietary guilds (folivore, frugivore and insectivore), following previous studies on strepsirrhine macroevolution [25, 27]. These categories reflect the predominant component of each species’ diet, rather than the level of dietary specialisation. A possible caveat with this approach is the inability to account for the role that gummivory plays in morphological variation, a dietary adaptation that has been associated with morphological specialisation in primates [32]. Phylogenetic relationships were reconstructed following Herrera and Dávalos [26], pruning their phylogeny to match our sample.

Brain morphology quantification

A dataset of 20 virtual cranial endocasts was obtained from Morphosource. This sample represents 20 species of 20 different genera (87% of generic diversity), covering the phylogenetic and ecological diversity within Strepsirrhini. Only adult specimens were included in our sample. Virtual endocasts were segmented using the 3D imaging software AVIZO® 9.1.1 software. Geometric morphometrics were implemented to quantify brain morphology, decomposing it into its size and shape components. A set of 30 anatomical landmarks were used to capture brain overall morphology (see S1 Table). Of these landmarks, 27 were developed by Bertrand et al. [33] and Ahrens [34]. An additional three were included to capture variation along the brainstem and the caudal-most aspect of the petrosal lobule [35]. In anticipation of future research, these landmarks were chosen based on their ability to capture shape variation across a morphologically diverse group, including other members of Euarchontoglires (i.e. Primates, Dermoptera, Scandentia, Rodentia, Lagomorpha). Landmarks were only placed on the left side of the endocast to avoid covariance between the same points on the left and right side of the endocast. All landmarking was performed in AVIZO® 9.1.1 using a WACOM Cintiq 21UX tablet. A generalised procrustes analyses (GPA) was implemented to control for methodological artifacts in the landmark data and to decompose brain morphology into its shape (isometry-free) and size (centroid size; CS) components. Linear regression models were used to assess the validity of using CS as a measure of brain size by quantifying the correspondence between it and BrM (R2 = 0.955, P< 0.001) and ECV (R2 = 0.962, P< 0.001).

Dental morphology quantification

Three dental topographic metrics were used to quantify dental morphology: Dirichlet Normal Energy (DNE) quantifies changes in curvature across the tooth crown as a proxy for tooth sharpness [36]; Orientation Patch Count Rotated (OPCR) estimates the number of distinct patches on the tooth crown that reflect the topographic complexity of the crown surface [37]; and Relief Index (RFI) measures tooth crown height as a ratio between the crown’s 3D and 2D surface areas [38]. Dental topographic metrics per specimen were gathered from one lower second molar, following previous protocols for primate dental topographic analysis [39]. Data for 146 specimens representing the 20 species sampled for brain morphological data was compiled from Fulwood et al. 2021a, 2021b [25, 27] and species averages were used to combine our brain and dental morphological datasets into a single dataset.

Allometry and phylogenetic structuring

To elucidate diet-driven ecomorphological patterns, we first quantified the effect of allometry and the presence of phylogenetic structuring in our data. Dental, CS and brain shape allometry were tested with a Procrustes regression (using log-transformed BdM), implemented in the procD.lm function in the R package Geomorph [40]. For CS and brain shape, the residuals of the Procrustes regression were used as allometry-corrected data in all further analyses. Phylogenetic structuring in dental morphology, CS and brain shape was assessed by estimating the multivariate K-statistic, using the physignal function in Geomorph with 10,000 iterations for significance testing [40]. The residuals of a phylogenetic linear regression of CS on log-transformed BdM were used as allometry-and-phylogeny-corrected CS and interpreted as a measure of relative brain size, using the procD.pgls function in Geomorph [40]. Given the significant phylogenetic signal and multidimensionality of our dental morphology and brain shape data, we performed a phylogenetic principal component analysis (pPCA) on the correlation matrix of each dataset to control for phylogenetic structuring and reduce the dimensions of our data, using the gm.prcomp function in Geomorph [40]. The first four principal components (each accounting for >10% of variation and in combination accounting for 57.32% of variation) were used for brain shape and the first three principal components (accounting for 100% of variation) were used for dental morphology in all further analyses. Due to our sample size, our per-guild evolutionary modelling analyses limited us to include the first four principal components of brain shape variation.

Diet-driven morphological variation

We studied whether diet explained patterns of brain and dental morphological variation after controlling for allometry and phylogenetic structuring. Differences in CS, brain shape and dental morphology across dietary guilds were assessed using phylogenetic ANOVAs with 10,000 iterations. Integration between dental and brain morphology was tested using two two-block partial least square analyses, one for CS and another for brain shape, using the two.b.pls function in Geomorph [40].

Morphological coevolution

We investigated differences in tempo and mode of brain and dental evolution across dietary guilds. To study differences in tempo, we first compared per-group net rates of morphological evolution across dietary guilds for each trait (dental morphology, brain shape and CS), under a Brownian motion model of evolution (BM) with the compare.evol.rates function in Geomorph [40] and 10,000 iterations. Pairwise comparisons of evolutionary rates between dietary guilds were also obtained from the previous analysis. We computed species-level morphological evolutionary rates using the phylogenetic ridge regression method developed in the R package RRphylo [41] for CS, brain shape and dental morphology, separately. We then assessed whether any of the dietary guilds represented a shift in evolutionary regimes, resulting in significantly higher or lower rates of evolution compared to the entire tree. To do this, we used the approach implemented in the search.shift function in RRphylo, using the sparse method and 1,000 iterations for significance testing [41]. We investigated the integration in evolutionary rates between brain and dental morphology using phylogenetic two-block partial least square analysis, accounting for the phylogenetic nonindependence in rates of evolution across species, as implemented in the phylo.integration function in Geomorph [40]. Finally, differences in the mode of evolution across dietary guilds were examined testing competing evolutionary models, using methods developed in the R package mvMORPH [42]. We pooled our data based on dietary guild and pruned our phylogeny to create subtrees for each guild. For each trait (dental morphology, CS and brain shape), we fitted three evolutionary models: BM that assumes a random-walk process, Early Burst (EB) assumes rapid early phenotypic diversification followed by a decline across time (i.e. adaptive radiation) and Ornstein-Uhlenbeck (OU) that assumes evolution under stabilising selection. Sample-size corrected Akaike information criteria (AICc) was used to select the best-supported models, with models with ΔAICc below 2 considered as supported. This procedure was performed for each dietary guild separately.

Results

Analyses of brain and dental morphological variation show a significant association in their macroevolutionary trajectories driven by dietary adaptations. Significant allometry was only found in brain shape and size, accounting for 11.55% and 93.25% of their variation respectively (Table 1). Statistically significant phylogenetic signal revealed a structuring in brain shape, size and dental morphological variation reflecting evolutionary relatedness (Table 2). After accounting for allometry and phylogenetic structuring, brain shape and dental morphology are significantly different between dietary guilds at P< 0.03 and relative brain size nearing significance at P = 0.07 (Fig 1 and Table 3). Folivores tended to cluster towards the negative end of pPC1, whereas insectivore species tended to group at the negative end of pPC3, with frugivore species showing the greatest overlap with other guilds (Fig 1A). The Aye-Aye occupied a unique subregion of morphospace, occupying the most positive region along pPC1-3. Comparing brain shape between species at opposite ends of each component revealed antero-posterior and dorsal-ventral distributions of shape variation along pPC1 and pPC2, respectively, whereas pPC3 concentrated most shape variation in the frontal lobe. Average relative brain size is markedly lower in folivores (lower than expected for their body mass), less variable in frugivores, and more variable in insectivores (Fig 1B). Dental sharpness (DNE) and crown height (RFI) are lower in frugivore species and higher in insectivore species, whereas average OPCR is higher in folivores and lower in frugivores (Fig 1C). Statistically significant morphological integration between dental morphology and both brain size (R2 = 0.493, P = 0.028) and shape (R2 = 0.795, P = 0.028) but not between brain size and shape (R2 = 0.896, P = 0.197) signal correlated brain-dental variation and decoupled brain size and shape variation.

Table 1

Procrustes linear regressions testing the effect of allometry in brain size and shape and dental morphology.

	Df	SS	MS	R2	F	Z	P
Brain shape	1	0.0267	0.0267	0.1156	2.3517	2.5053	0.0049
Brain size	1	1.8043	1.8043	0.9325	248.8100	6.8328	0.0001
Dental morphology	1	0.1820	0.1820	0.0716	1.3876	0.6537	0.2739

Table 2

Tests of phylogenetic signal in brain size and shape and dental morphology.

	K	Z	P
Relative brain size	0.743	1.929	0.026
Brain shape	0.589	2.786	0.002
Dental morphology	0.953	2.191	0.008

Fig 1

Patterns of strepsirrhine phenotypic variation across dietary guilds.

Phylomorphospace of allometry-controlled brain shape based on the phylogenetic principal component analysis (A); boxplots of allometry-controlled centroid size (relative brain size) across dietary guilds (C); boxplots of dental morphological variation across dietary guilds (C), representing Dirichlet Normal Energy (DNE; bottom left), Orientation Patch Count Rotated (OPCR; bottom middle) and Relief Index (RFI; bottom right). Endocast heatmaps exemplify differences in brain shape between species occupying opposite ends of each principal component. Heatmaps were obtained by warping the endocast of the modern taxon closest to the inferred average morphology (Eulemur fulvus) based on the Procrustes coordinates of species on opposite ends of a principal component and estimating the distance between them.

Table 3

Procrustes ANOVAs testing for differences in allometry- and phylogeny-corrected brain size and shape and dental morphology across dietary guilds.

	Df	SS	MS	R2	F	Z	P
Brain shape	2	0.073	0.036	0.160	1.621	2.018	0.023
Relative brain size	2	0.035	0.018	0.264	3.055	1.430	0.075
Dental morphology	2	1.109	0.555	0.265	3.063	1.913	0.028

Per-branch rates of morphological evolution revealed similar patterns in dental morphology and brain shape variation across species, while relative brain size evolution follows an independent pattern (Fig 2A–2C). Interestingly, D. madagascariensis was the only species to have high evolutionary rates in all three traits. Frugivore species have lower evolutionary rates across all three traits, showing statistically significant differences in rates of brain shape evolution (Fig 3 and S2 Table). Comparisons of per-group evolution rates showed frugivores have the lowest net morphological evolution rates in all three traits, albeit only statistically significantly different in brain shape (Fig 3A and 3B, S3 and S4 Tables). Folivores have the highest per-guild and per-species rates of brain size and shape evolution, whereas insectivores have the highest rates of dental morphological evolution (although none was statistically significantly different, see S2–S4 Tables). Significant integration in rates of brain shape and dental morphological evolution (S5 Table), after controlling for the phylogenetic nonindependence in evolutionary rates across species (Table 4), supports the coevolution of brain and dental phenotypes driven by diet (Fig 3C).

Fig 2

Per-species rates of phenotypic evolution based on phylogenetic ridge regressions for brain shape (A), brain size (B) and dental morphology (C).

Symbols represent dietary guilds.

Fig 3

Comparisons of rates of brain shape evolution across dietary guilds.

Pairwise comparisons of per-guild evolutionary rates in brain shape (A); differences in per-species rates of evolution averaged by guild and compared to the average rate for all strepsirrhines combined (B); partial least square analysis of integration between per-species rates of evolution of brain shape and dental morphology (C). * Represents statistical significance at P< 0.05.

Table 4

Phylogenetic two-block least square analyses for integration in evolutionary rates between brain size and shape and dental morphology.

	R2	Z	P
Brain shape_ Relative brain size	0.107	0.465	0.665
Brain shape_Dental morphology	0.493	1.968	0.028
Brain size_Dental morphology	0.332	1.405	0.156

Best-supported models of brain and dental morphological evolution differ between dietary guilds (Table 5). Brain shape evolved under stabilising selection (OU model) in folivores and insectivores and following a random-walk (BM) process in frugivores, brain size evolution followed the opposite pattern folivores and insectivores, whereas all three traits evolved under no directional selection (BM) in frugivores. Dental morphology evolved under stabilising selection in folivores and following a BM process in frugivores and insectivores.

Table 5

Model fitting testing competing evolutionary hypotheses of phenotypic diversification across strepsirrhine dietary guilds.

Three evolutionary models were tested: Brownian motion (BM), early burst (EB) and Ornstein-Uhlenbeck (OU). Models with ΔAICc lower than 2 were inferred as best supported.

Variable	Guild	Model	LogLik	AICc	ΔAICc
Relative brain size	Insectivore	BM	5.116	-2.232	0.000
		EB	5.116	7.768	10.000
		OU	5.730	6.541	8.773
	Folivore	BM	5.822	-1.645	0.000
		EB	5.822	18.355	20.000
		OU	5.829	18.342	19.987
	Frugivore	BM	13.640	-21.279	0.000
		EB	13.640	-16.479	4.800
		OU	15.681	-20.562	0.717
Brain shape	Insectivore	BM	45.553	-16.438	1234.304
		EB	45.153	-0.307	1250.435
		OU	49.371	-1250.742	0.000
	Folivore	BM	49.056	13.888	311.249
		EB	48.708	52.584	349.945
		OU	52.680	-297.361	0.000
	Frugivore	BM	82.048	-116.096	0.000
		EB	80.554	-107.108	8.988
		OU	87.956	-18.821	97.275
Dental morphology	Insectivore	BM	-19.597	79.693	0.000
		EB	-19.597	90.622	10.929
		OU	-16.810	303.621	223.927
	Folivore	BM	-5.281	64.563	508.424
		EB	-5.281	85.563	529.424
		OU	-3.069	-443.861	0.000
	Frugivore	BM	-9.070	46.728	0.000
		EB	-9.070	51.890	5.162
		OU	-2.992	79.619	32.891

Log-likelihood (LogLik), sample-size corrected Akaike information criterion (AICc) and relative fit (ΔAICc) are shown.

Discussion

Contrary to our predictions, our results show a significant effect of diet on the phenotypic coevolution of the brain and dental morphology of strepsirrhines. Adaptations to both the brain and dentition were crucial factors during the evolution of primates, enabling their ecological diversification [15, 18, 19]. From their early divergence from haplorhines, strepsirrhines evolved remarkably high morphological variability associated with their ecological diversification. Contrary to previous primate macroevolutionary studies that depict strepsirrhine brain evolution as a period of relative stasis [13, 43], our results show a complex multifaceted evolutionary process. After controlling for allometry and phylogenetic structuring, our results showed a significant effect of diet on dental and brain morphology, and a strong integration in the variation and evolutionary trajectories of dental morphology and brain shape, but not brain size. This is in stark contrast with multiple previous studies that have investigated brain evolution only in terms of size [4, 8, 9], and highlights the potential of studying other phenotypic dimensions of brain evolution [3, 14, 15]. Combined, our results provide clear evidence for the effect of diet in the variation and integration of brain and dental morphology during strepsirrhine diversification. Furthermore, our results emphasise the differential effect of factors like diet and sociality have for the evolution of the brain across multiple primate groups [5, 10], highlighting the importance of studying macroevolutionary processes at multiple phylogenetic scales. Given the unique evolutionary trajectory of the Aye-Aye in our results, we replicated our analyses while excluding it to test the sensitivity of our results. All of the notable conclusions hold after removing the Aye-Aye (see S1 File).

Decoupled patterns of brain size and shape variation and evolution in our results support previous findings in hominins [19], platyrrhines [44] and catarrhines [45] that also found independent patterns of brain size and shape variation. Moreover, our results of brain shape allometry suggest that changes in body mass only explain a fraction of shape variation, as reported in platyrrhine primates [15]. Despite this decoupling of brain size and shape, both traits differed across guilds, indicating independent patterns of adaptation to dietary differences. Significant brain shape allometry in our results contrast with a previous report of non-allometric variation in brain shape in strepsirrhines [13]. The fact that differences in relative brain size between dietary guilds were only nearing significance might be a result of unintended loss of ecological signal after correcting for phylogenetic relatedness, as ecological adaptations also tend to exhibit phylogenetic patterning in primates [46]. Our results of smaller relative brain size in folivores agree with previous findings in strepsirrhine primates [8]. Significant differences in brain shape between guilds also suggest dissimilar adaptations across brain regions depending on different cognitive specialisations, highlighting the organisational complexity of the brain [3, 14]. In strepsirrhines, differences in the relative size of olfactory and visual sensory brain regions have been identified between species with different diets and activity periods (e.g. olfactory structures are enlarged in frugivores; [14]). Compared to other primates, strepsirrhines have relatively enlarged spatial cognition brain areas, which has been associated to their foraging behaviour [14]. Visualisation of brain shape variation between the mean shape of our complete sample and the mean shapes of each dietary guild reveal guild-specific patterns, the insectivore mean shape varying mostly in the dorsal-most aspect of the neocortex, the folivore mean shape primarily in the frontal lobe and the frugivore mean shape varying in a less well-defined pattern across the brain (S1 Fig). While this visualization method identifies shape variation in specific regions of the neocortex (i.e. frontal lobe), additional testing is required to better isolate variation within these subregions with an expanded suite of landmarks specifically tailored to this goal, analyses that are beyond the scope of this study. The present study supports an increasing body of research that points to the need for a better understanding of the modular organisation of the brain [3, 14, 15] and thereby expand our current understanding of primate brain evolution (mainly focused in size). Future studies could apply geometric morphometrics to elucidate integration patterns across brain regions in response to ecological specialisations.

Dental adaptations linked to dietary specialisations in strepsirrhines in our study followed general patterns of dental dietary ecomorphology previously reported in primates [23, 25, 47, 48]. Insectivorous strepsirrhines had higher DNE values, indicating sharper molars adapted to crush and fragment invertebrates’ exoskeleton [23]. Similarly, taller tooth crowns (represented by high RFI values) in insectivorous species reflected previous findings indicating decreasing crown height along an animalivory-herbivory gradient in prosimians and platyrrhine primates [23]. Finally, topographic complexity of the molar crown did not reveal a clear pattern discriminating dietary specialisations, although mean OPCR was higher in folivores and lower in frugivores. Ambiguous differentiation of dietary guilds based on OPCR has previously been reported in prosimians, especially after controlling for phylogenetic relatedness [23, 47] Unexpectedly, the highest variation in OPCR was found in folivorous species, which could reflect the importance of specialisations to different folivore niches during the diversification of Malagasy lemurs [49]. Significant integration between dental morphology and brain size and shape suggests that, despite their assumed independence, dietary specialisations represented a major factor canalising their variation to act as a functional unit. Studying the association between brain and dental size in primates, prosimians showed a unique pattern of positive correlation between these two factors and with increased dietary quality [18]. Our results showing positive integration in brain and dental morphology provide evidence that the strepsirrhine brain-dental morphofunctional association extends beyond similarities in size [18].

Our analyses of the mode and tempo of brain and dental phenotypic evolution revealed diet played an important role throughout strepsirrhine evolution, reflecting the general pattern found for primate brain size evolution [4, 8, 9, 14]. However, we found a consistent pattern of lower evolutionary rates across brain and dental morphological traits associated with frugivory, suggesting that frugivory might not have promoted phenotypic diversification in strepsirrhines. Palaeoecological reconstructions have hypothesised that the strepsirrhine common ancestor was probably a frugivore, from which the clade diversified to specialise in a variety of dietary niches [24, 27]. Lower rates of brain and dental evolution in frugivores in our results would be consistent with frugivory as the ancestral state for strepsirrhines. Moreover, our results of higher evolutionary rates in folivores indicate unexpected phenotypic changes to adapt to folivorous niches. Malagasy colonisation by strepsirrhine primates has been associated with a diversification of folivore niches, linked to significant molecular adaptations to occupy non-overlapping niches and avoid competition [49]. Our results showing accelerated phenotypic changes in folivores can be linked to the folivorous diversification of malagasy lemurs. We hypothesise that previous macroevolutionary studies on the evolution of primate and mammalian brain failed to uncover this unique pattern of folivory-based phenotypic evolution in strepsirrhines because they were being swamped by higher level patterns of change. Evolutionary model fitting revealed that the two most-integrated traits (brain shape and dental morphology) followed the same trajectory, but that trajectory differed across guilds. Brain shape and dental morphology evolved under stabilising selection in folivores and following a BM process in frugivores, signalling the selective pressures acting during the phenotypic evolution of folivorous strepsirrhines and the ancestral frugivorous state in Strepsirrhini [25, 28, 49]. Our results suggest that adapting to insectivory correlated with directional selection in brain shape, the only guild with such a pattern. We hypothesise that unique functional demands associated to insectivores’ foraging behaviour are linked to adaptations in specialised brain regions (possibly visual and auditory signal processing regions), rather than an adaptation in whole-brain size or dental morphology [14].

In conclusion, our study provides further evidence of the role diet played during the phenotypic evolution of strepsirrhine primates, revealing clade-specific processes of brain and dental adaptations to folivory and insectivory in one of the earliest primate radiations. In contrast to previous studies in hominins, we show a strong integration between brain and dental phenotypic change even after controlling for allometry and phylogenetic relatedness. Surprisingly, we found an extreme pattern of phenotypic specialisation and accelerated evolution in the Aye-Aye, suggesting its colonisation of Madagascar represented a unique evolutionary event, with a pattern and magnitude of change distinct from those observed in other parts of the strepsirrhine tree that merits further studies [29, 30, 50]. It is possible that our findings of unusual morphological evolution in the Aye-Aye reflect both the relatively recent divergence between the Aye-Aye and its closest relative (Daubentonia robustus) during the Pliocene [26], and the extreme morphological convergence between the Aye-Aye and sciurid rodents [30]. Finally, our study reveals the importance of brain shape evolution to understand the diversification of strepsirrhines, as well as primates and vertebrates more generally, shedding light on evolutionary processes otherwise overlooked by studying brain size alone. Future studies should explore the role that gummivory had during the ecomorphological evolution of Strepsirrhini and mammals in general.

Description of anatomical landmarks.

Description of anatomical landmarks used to capture brain morphology and analogous landmarks from Bertrand et al. [33] and Ahrens [34].

(DOCX)

Click here for additional data file.

Test for differences in evol rates.

Statistical test for differences in per-species evolutionary rates across dietary guilds and pairwise comparisons.

(DOCX)

Click here for additional data file.

Test for differences in per-guild evol rates.

Statistical test for differences in per-guild evolutionary rates across dietary guilds. Net evolutionary rates are provided per guild for each trait.

(DOCX)

Click here for additional data file.

Pairwise tests of evol rates differences.

Significance values for pairwise statistical tests of differences in per-guild evolutionary rates across traits.

(DOCX)

Click here for additional data file.

Phylogenetic signal in evol rates.

Test of phylogenetic signal in per-species evolutionary rates.

(DOCX)

Click here for additional data file.

Visualisation of guild-specific brain shape adaptations.

Visualisation of brain shape (Procrustes coordinates) variation between the mean shape of our complete sample and the mean shapes of each dietary guild reveal guild-specific patterns.

(EPS)

Click here for additional data file.

Set of replicated results without the Aye-Aye.

(DOCX)

Click here for additional data file.

Raw Landmark data.

(TXT)

Click here for additional data file.

Raw dental topographic data.

(CSV)

Click here for additional data file.

R code to reproduce analyses.

(R)

Click here for additional data file.

11 Apr 2022

Submitted filename: Review for Lopez-Aguirre.pdf

Click here for additional data file.

Submitted filename: PONE-D-22-04796_reviewer.pdf

Click here for additional data file.

2 May 2022

ONE-D-22-04796

Diet drove brain and dental morphological coevolution in strepsirrhine primates

PLOS ONE

Dear Dr. Lopez-Aguirre,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

First of all, I want to express to you that I am sorry for the delay in making this return on your work. Finally, we have received the opinions of three reviewers who have made some comments and suggestions, which is why I have decided to request minor revisions.

Basically, I ask you to pay attention mainly to reviewer 2's methodological criticisms and to respond, as usual, whether or not you accept each of the comments and suggestions of the three reviewers. If you are unable to do so or disagree, please justify this in your reply to this editor.

Response. We thank the editors and reviewers for their comments and suggestions. Some changes to the manuscript and analyses were made in order to address both reviewers’ comments. Please see responses to specific comments below.

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

Response: Done. We reviewed the guidelines and edited the manuscript accordingly.

2. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

Response: Done. We have revised our data availability statement to specify that all data and code will be available within the supporting information files.

3. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Response: Done.

Review Comments to the Author

Reviewer #1: I appreciated the opportunity to review this manuscript on the coevolution of dental morphology and brain shape in strepsirrhine primates. In this study, the authors examined brain and dental evolution in the context of different dietary guilds (frugivory, folivory, insectivory). They found that brain shape and dental morphology evolved within dietary guilds at similar rates and patterns, suggesting a link between diet and the coevolution of these morphological features. The analyses are appropriately rigorous and the authors’ conclusions are supported by their data (some exceptions noted below, which require further clarification). This study fills a gap in the existing literature on brain/dental evolution in primates, providing a more specific look at this suborder as well as highlighting the importance of examining shape variables in addition to size. In general, I believe this work is suitable for publication in PLOS ONE, pending some revisions. Below, I list some major and minor suggestions for improving the manuscript. I would be happy to review a revised manuscript.

Response: We appreciate the reviewer’s positive comments and constructive suggestions on the manuscript.

Major:

Page 9: “Species cluster in different subregions of brain morphospace based on dietary guild, with some overlap across guilds where more generalist species group together (Fig. 1).” This feels like an overstatement, based on the considerable overlap in the PCA. Which are the generalist species? They aren’t discussed elsewhere in the manuscript, even though it seems that this could be important context for the dietary guilds. Also, you may wish to address the insectivore outlier in the PCA (Aye-Aye?).

It would be great to include an interpretation of the principal components used for brain shape (Fig. 1); the loadings for these components could be included in a supplementary data file.

Response: Done. We have included heatmaps of endocasts comparing the shape of the two species on opposite ends of each principal component, showing where shape change is concentrated. With this, we have replaced our description of the PCA in the Results with a more accurate and informative section. See page 9. We avoid discussing the loadings of these components as they are rotated based on phylogenetic relatedness, making it impossible to infer detailed morphological changes from them.

Relatedly, on pages 13-14, you address the potential ecological factors driving the evolution of brain shape/cognitive specialization. What kinds of variation in brain shape are associated with the different dietary guilds (in this study, not just in the existing literature)? While I understand that an in-depth analysis of the evolution of brain regions is beyond the scope of this paper, some more detail is warranted here, especially since the following paragraph provides a finer-grained interpretation of DNE, RFI and OPCR values in dental evolution across the three guilds.

Response: Done. We have included a supplementary figure visualising shape variation between the mean shape of our complete sample and the mean shape of each dietary guild, and briefly discussed guild-specific patterns in the discussion. As the reviewer said, an in-depth analysis of guild-specific brain adaptations is beyond the scope of the current study, so we’ve tried to provide additional information without venturing into speculative arguments. See pages 14-15 and S1 Fig.

Page 11: “Brain shape evolved under stabilising selection (OU model) in folivores and insectivores and following a random-walk (BM) process in frugivores, whereas brain size evolution followed the opposite pattern across guilds.” This statement is confusing in light of Table 5, which indicates that both the BM and OU models have low delta AICc values for frugivore brain size. In fact, the BM model has a lower value, suggesting that brain size does not necessarily follow the opposite pattern as shape across all guilds.

Response. We thank the reviewer for highlighting this inconsistency. We have amended this sentence to better reflect the different patterns across guilds and how all three traits seemed to have evolved under no directional selection (BM) in frugivores. See page 11.

On page 15, you provide the paleoecological context for evolutionary rates in folivores and frugivores; this interpretation/context is missing for rates of insectivore evolution.

Response. Done: We have included two additional sentences in page 16 specifically discussing our findings with respect to insectivory and strepsirrhine evolution. We argue that our results indicate insectivory followed a unique evolutionary trajectory where it only correlated with directional selection in brain shape. Moreover, we hypothesise that unique functional demands associated with insectivores’ foraging behaviour are linked to adaptations in specialised brain regions (possibly visual and auditory signal processing regions), rather than an adaptation in whole-brain size or dental morphology.

Minor:

The introduction would benefit from a smoother transition between the paragraphs on page 4 (from dental topographic analysis to evolutionary history of strepsirrhines). This could be an opportunity to justify your focus on this taxonomic group; why is it a good study system to address the aforementioned questions? The answer is implicit in the following paragraph, but could be better articulated here.

Response. Done: We have included an additional sentence in-between the two paragraphs in question to smooth the transition of topics and better highlight the importance of focusing on studying strepsirrihine primates. See page 4.

Gummivory is mentioned once in the introduction (page 4) but is not included in the analyses or addressed later in the manuscript. Since you mention it earlier, it might be worth acknowledging and justifying its absence in the study.

Response: Done. Following suggestions made by reviewers 1 and 3, we have better clarified our dietary categories and clearly stated the caveat in our classification and the importance of exploring the effect of gummivory on the ecomorphological evolution of Strepsirrhini. See pages 6 and 16.

I assume that all the data were collected from adult individuals, as age impacts brain/dental size and shape. This should be explicitly stated.

Response: Done. We have added additional information clarifying that all specimens analysed were adults. See page 6.

Page 7: What is the justification for including the first four principal components (and not more or fewer) as a proxy for brain shape?

Response: Done. We have included additional information in the methods to better explain our reasoning for this (see page 7). Our per-guild sample size limited our capacity to include more principal components in the evolutionary modelling analyses, so we decided to include: 1) all the components that each explained at least 10% of variation and 2) as many as the modelling analyses could take.

I strongly suggest labeling different panels within each figure with A, B, C, etc., so that you can clearly refer to specific elements of each figure in the text.

Response: Done. We have labelled different panels within each figure and amended the results and legends accordingly.

I was glad to see you address the Aye-Aye’s extreme accelerated evolution at the very end of the discussion (page 16). This is an interesting result, and an additional sentence speculating about the possible ecological reasons for this pattern would be warranted here.

Response: Done. We have included an additional sentence interpreting our results as possible evidence of the relatively recent divergence between the Aye-Aye and its closest relative (D. robustus) in the Pliocene, and the extreme morphological convergence between the Aye-Aye and sciurid rodents. See page 16.

Reviewer #2: Do I think this paper should be published? No, not really. I will enumerate the many reasons below. However, rather than trying to block this paper from making it into print, I am going to recommend—counter to my own opinion of the work—that this paper is accepted, I would say without revision, but I leave it up to the authors to revise it as they please and resubmit. I promise to accept without further revision, or to work with the authors until either there is something publishable (which I doubt) or they feel they’ve made enough of an effort that this deserves to be in print (its really up to them). First, I will explain why I am taking this counter-intuitive tact. Then I will explain why this paper is a misadventure in P-Hacking, in the attached document.

Response. The reviewer has provided a detailed review, providing in-depth opinions on a variety of topics and raising a myriad of issues. Unfortunately, the tone the reviewer used when writing the revision muddies the criticisms they are trying to convey, which led the reviewer to extensively discuss things ranging from the peer-review process generally and scientific publishing as a whole, to questioning our work ethic and desire to do high-quality science. Confusingly, the reviewer concludes promising “to accept without further revision” while at the same time promising “to work with the authors until … there is something publishable (which I doubt)”. Based on this comment, we fail to see how the reviewer is interested in engaging in a constructive and professional scientific exchange of ideas. Despite these differences, we have gone through the reviewer’s comments and will provide a point-by-point response to the main issues. First, we respectfully disagree with the reviewer’s overall assessment of our study, but especially with the tone and the intentions that seem to have motivated this review. We might share some of the reviewer’s concerns in terms of the peer-review process and scientific publishing, but contrary to the reviewer we don’t see this as a productive setting to discuss this. In terms of the specific criticisms to our study, we are addressing the main points:

1) P-Hacking: The reviewer argues that our study is “a misadventure in P-Hacking”. The reviewer employs a series of interpretations of aspects of our study to justify why they’re certain that we engaged in P-hacking in order to find “significant results”. Unsurprisingly, we strongly disagree with the reviewer on this point for several reasons; 1) The reviewer states that “the premise and conclusions of this paper are not controversial”, which in our opinion seems counterintuitive to the need for P-Hacking. If our results are so underwhelming and predictable, why would we need to doctor our data or statistical analyses to find an already predictable and expected pattern? 2) We clearly stated our reasoning and expectations for our results in the manuscript, and as we clearly stated in our discussion, our results were contrary to our expectations in some ways. We followed a structured and sequential line of observation, questioning and testing, starting with a premise and when our results didn’t align with it, we reoriented our discussion to try and explain the unexpected patterns we found. Our counterargument would be, how could we have engaged in P-hacking when some of our results go against our expectations and predictions? 3) The reviewer also states that we implemented a “strange and unnecessary over-application of various analytic tools”, while at the same time criticising the ability of individual analyses to test our hypotheses. We are aware of the different limitations some of our analyses have, which is precisely the reason why we decided to use a combination of tests to thoroughly assess our data. Rather than relying on a single test with limited capacity to investigate our data, we strove to compensate for the limitations that come with each test.

2) The effect of the Aye-Aye: The reviewer repeatedly argued in multiple occasions that our results were driven by the unusual biology of the Aye-Aye and that without it, all our findings wouldn’t hold. The reviewer also argues that we should just not analyse the Aye-Aye because it’s a “biological outlier”. We respectfully disagree with the premise that one can simply deem a species a “biological outlier” and ignore it. Ignoring the presence of species with unique adaptations that represent rare evolutionary processes artificially homogenises and standardises nature. Having said that, we agree with the reviewer that it is important to test the sensitivity of our results to the presence of the Aye-Aye, so we have included an additional set of analyses excluding the Aye-Aye. The results of these additional analyses show that our results are not driven by this one species’ unusual evolutionary trajectory (see S1 File with five supplementary tables of results).

3) Allometry: The reviewer disagrees with our usage of residuals as an allometry-corrected trait. 1) Contrary to the reviewer’s argument, not all phylogenetic comparative methods allow the inclusion of confounding factors (e.g. comparing evolutionary rates across groups while accounting for allometry), this is the reason why it is common practice to use the residuals in these kind of analyses, and 2) The reviewer’s assumption that allometric trajectories have to vary across taxa (which is true but not a universal pattern, as has been shown across several groups) and that this renders the use of residuals non-informative would be relevant to a different research question altogether, i.e., how allometric trajectories may vary across taxa. That is not the focus of this study, so we argue that the use of the residuals of a single allometric trajectory for our sample is valid, as it is only used to test and account for the effect of size in our sample.

4) Accounting for phylogenetic signal: The reviewer criticises our usage of the components of a phyloPCA to obtain data that accounts for phylogenetic relatedness for its use in subsequent phylogenetic analyses. Since what we are testing here is the effect of a single ecological trait (diet) on the evolution of brain and dental morphology, not accounting for the possible effect of phylogenetic relatedness on morphological traits would make it impossible to truly discern the effect of diet. Moreover, controlling for phylogenetic relatedness in the morphological data doesn’t erase the phylogenetic information in our analyses, because that is explicitly incorporated by way of the phylogeny that is used as a scaffold for all analyses. Additionally, these methods cannot account for the possible phylogenetic structuring in our morphological data while at the same time reconstructing their evolution based on biotic or abiotic traits (explicitly untangling whether morphological variation is linked to evolutionary relatedness or ecological factors). Therefore, it is common practice to use ordination analyses to reduce the dimensionality of the data while accounting for relatedness, in order to test the effect of an ecological or biological factor like diet. Not using our approach would limit our results to reconstructing macroevolutionary trajectories, without investigating mechanistic models like the effect of diet on such trajectories.

5) Data and code: Contrary to the reviewer’s assertion that we are “withholding” the data, we clearly stated in our data availability statement that all raw data and code will be made freely accessible. In this revision we are including our raw data and code in the supplements.

Reviewer #3: The evolution of primate brain was always of high interest and a matter of debate. Several works have been presented on the subject sometimes differing depending on the primate group analyzed. The present manuscript relating diet to brain and dental morphology coevolution is of remarkable interest due to the diversity and adaptations observed in strepsirrhines, and added important insights to consider for future research on this topic.

The methodology is adequate and was perfectly applied by the authors. Something that may be explained in more detail is the selection of three categories of diet, since it is somewhat difficult to precisely define some categories as frugivorous-folivorous, and that insectivores and folivores may differ in size to be classified. Also, gummivory is a critical category that may lead to unique adaptations. It is suggested to specify why the authors selected those three categories despite others mixed or intermediate to develop the work, although the results were satisfactory by applying the methods.

Among the main results, brain and dental morphology are integrated meaning that diet has major influence to explain their adaptations, differing from some previous studies, and it is notable that especially brain shape and dental morphology are the most integrated traits. These results allow to explore integrations in other primate groups.

Response: We appreciate the reviewer’s positive comments and constructive suggestions on the manuscript.

Upon some explanations detailed in the text, I recommend publication of this work.

Response: Done. Following suggestions made by reviewers 1 and 3, we have better clarified our dietary categories and the importance of exploring the effect of gummivory on the ecomorphological evolution of Strepsirrhini. See pages 6 and 16.

Submitted filename: RebuttalLetter _F.docx

Click here for additional data file.

13 May 2022

Diet drove brain and dental morphological coevolution in strepsirrhine primates

PONE-D-22-04796R1

Dear Dr. Camilo López-Aguirre,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Claudia Patricia Tambussi, Ph.D.

Academic Editor

PLOS ONE

Additional Editor Comments

I appreciate that you have evaluated all the opinions expressed by the three reviewers, focusing on those specifically related to the contents of this work. The main questions raised by them, especially reviewer 1 and 3 have been satisfactorily considered in this new version of the manuscript.

27 May 2022

PONE-D-22-04796R1

Diet drove brain and dental morphological coevolution in strepsirrhine primates

Dear Dr. Lopez-Aguirre:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Claudia Patricia Tambussi

Academic Editor

PLOS ONE