Overall, six meta-regressions were calculated (ambilaterality, leftward lateralization, and rightward lateralization, each with adult brain mass in g and neuron number as predictors). Data were collected from 28 different species (see Table 1). Brain size data were available for all 28 species, neuron number data for only 17 species.
Table 1
Studies on hemispheric asymmetries, brain size, and neuron numbers that were included in the meta-regressions. Numbers indicate the relevant references in the reference list.
Species | Study asymmetry | Study brain size | Study neuron number |
Domestic sheep (Ovis aries) | 32 | 27 | - |
Domestic pig (Sus scrofa domesticus) | 33 | 27 | 27 |
Domestic dog (Canis familiaris) | 10 | 27 | 27 |
Domestic cat (Felis catus) | 10 | 27 | 27 |
Red Kangaroo (Macropus rufus) | 34 | 27 | - |
Eastern Grey Kangaroo (Macropus giganteus) | 34 | 35 | - |
Red-necked Wallaby (Notamacropus rufogriseus) | 36 | 37 | 37 |
Goodfellow's tree-kangaroo (Dendrolagus goodfellowi) | 34 | 37 | 37 |
Common Marmoset (Callithrix jacchus) | 38 | 27 | 27 |
Gorilla (Gorilla gorilla) | 39 | 27 | 40 |
Orang utan (Pongo pygmaeus) | 39 | 27 | 40 |
Chimpanzee (Pan troglodytes) | 39 | 27 | - |
Bonobo (Pan paniscus) | 39 | 27 | - |
Ring-tailed Lemur (Lemur catta) | 41 | 27 | - |
Long-tailed Macaque (Macaca fascicularis) | 42 | 27 | 27 |
Rhesus macaque (Macaca mulatta) | 42 | 35 | 43 |
Southern pig-tailed macaque (Macaca nemestrina) | 42 | 44 | - |
Eastern Gray Squirrel (Sciurus carolinensis) | 45 | 27 | 27 |
Sugar glider (Petaurus breviceps) | 46 | 47 | - |
Grey short-tailed opossum (Monodelphis domestica) | 46 | 47 | - |
Squirrel Monkey (Saimiri sciureus) | 48 | 43 | 43 |
Guinea baboon (Papio papio) | 49 | 35 | - |
Olive baboon (Papio anubis) | 50 | 51 | - |
Gray Mouse Lemur (Microcebus murinus) | 52 | 43 | 43 |
Tufted Capuchin (Cebus apella) | 53 | 43 | 43 |
House Mouse (Mus musculus) | 9 | 27 | 27 |
Rat (Rattus norvegicus) | 9 | 27 | 27 |
Human (Homo sapiens) | 8 | 27 | 27 |
Insert Table 1 here
For limb preferences and adult brain mass, the ambilaterality meta-analysis (see Fig. 1 for forest plot) revealed an overall proportion of ambilateral limb preferences across species that was 0.30 (95% confidence interval: 0.22 to 0.39). Thus, across species 30% of animals show an ambilateral preference. Significant heterogeneity across studies was detected (Q(27) = 2325.75; p < 0.001). Meta-regression with brain mass as predictor did not reach significance (F(1,26) = 0.4122; p = 0.53). This suggests that brain mass is not associated with the number of ambilateral individuals in a species.
Insert Fig. 1 here
The meta-analysis for leftward lateralization (see Fig. 2 for forest plot) revealed an overall proportion of rightward limb preferences across species that was 0.31 (95% confidence interval: 0.23 to 0.40). Thus, across species 31% of animals show a leftward preference. Significant heterogeneity across studies was detected (Q(27) = 9429.12; p < 0.001). Meta-regression with brain mass as predictor reached significance (F(1,26) = 6.77; p < 0.05), with a negative t-value of t=-2.60 for the predictor brain mass. This suggests that higher brain mass is associated with a lower number of individuals with a leftward preference in a species.
Insert Fig. 2 here
The meta-analysis for rightward lateralization (see Fig. 3 for forest plot) revealed an overall proportion of rightward limb preferences across species that was 0.33 (95% confidence interval: 0.24 to 0.44). Thus, across species 33% of animals show a rightward preference. Significant heterogeneity across studies was detected (Q(27) = 8861.70; p < 0.001). Meta-regression with brain mass as predictor reached significance (F(1,26) = 4.42; p < 0.05), with a positive t-value of t = 2.10 for the predictor brain mass. This suggests that higher brain mass is associated with a higher number of individuals with a rightward preference in a species.
Insert Fig. 3 here
For the meta-analyses with neuron number as predictor in the meta-regression we do not present the forest plots, as this information is already included in the Figs. 1–3. The ambilaterality meta-analysis revealed an overall proportion of ambilateral limb preferences across species that was 0.30 (95% confidence interval: 0.19 to 0.44). Thus, across species 30% of animals showed an ambilateral preference. Significant heterogeneity across studies was detected (Q(16) = 1900.33; p < 0.001). Meta-regression did not reach significance (F(1,15) = 0.96; p = 0.34). This suggests that the number of neurons in the brain is not associated with the number of ambilateral individuals in a species.
The meta-analysis for leftward lateralization revealed an overall proportion of rightward limb preferences across species that was 0.28 (95% confidence interval: 0.19 to 0.38). Thus, across species 28% of animals showed a leftward preference. Significant heterogeneity across studies was detected (Q(16) = 9091.78; p < 0.001). Meta-regression reached significance (F(1,15) = 5.07; p < 0.05), with a negative t-value of t=-2.25 for the predictor number of neurons in the brain. This suggests that a higher number of neurons in the brain is associated with a lower number of individuals with a leftward preference in a species.
The meta-analysis for rightward lateralization revealed an overall proportion of rightward limb preferences across species that was 0.34 (95% confidence interval: 0.21 to 0.49). Thus, across species 34% of animals show a rightward preference. Significant heterogeneity across studies was detected (Q(16) = 8216.07; p < 0.001). Meta-regression reached significance (F(1,15) = 4.69; p < 0.05), with a positive t-value of t = 2.17 for the predictor number of neurons in the brain. This suggests that a higher number of neurons in the brain is associated with a higher number of individuals with a rightward preference in a species.
Taken together, the results for adult brain mass and neuron number in the brain paralleled each other completely. For both, the predictor failed to show an association with ambilaterality but showed a significant positive association with the prevalence of rightward preferences in a species and a significant negative association with the prevalence of leftward preferences in a species. This finding is only partly in line with the Ringo hypothesis 12 and the preregistered hypotheses of the present study. Two of the six preregistered hypotheses were confirmed. Brain mass and neuron number were statistically significant predictors for the number of animals with rightward lateralization and the directionality of the effect was positive (e.g., species with larger brains showed more rightward lateralization). The statistical tests for the preregistered hypotheses for ambilaterality, however, did not reach significance. For leftward lateralization we found effects that were significant but opposite to what was predicted in the preregistered hypotheses. This suggests that interhemispheric conduction delay may play a role in the evolution of functional hemispheric asymmetry but may not be as central as suggested by the Ringo hypothesis 12.
The Ringo hypothesis predicts a general shift away from ambilaterality toward laterality in larger-brained species but makes no prediction on the direction of laterality. In contrast to that, our findings suggest a specific shift toward rightward limb preferences in larger-brained species and a reduced number of leftward limb preferences. This is most evident in humans, with their distinct 90:10 distribution for right-handedness and left-handedness 8. While the methodology of the present meta-regression study does not allow for causal inferences, it is evident that other factors than interhemispheric conduction delay need to be considered in the context of the evolution of hemispheric asymmetries. It could be speculated that sociality may be a factor that also plays a role, as it has been implied in both the evolution of brain size 18 and the evolution of hemispheric asymmetries19,20. One leading theoretical account for the evolution of population level hemispheric asymmetries within a species suggests that population level asymmetries emerge as an evolutionary stable strategy when organisms need to coordinate their behavior with other asymmetrically behaving individuals 21. This implies that in particular social species should show population level asymmetries towards one side, an idea that is supported by empirical evidence in insects 19. In this study it was shown that the social honeybee shows hemispheric asymmetries on the behavioral and electrophysiological level, while the non-social mason bee does not. Moreover, the results of a study on handedness and learning how to fold asymmetric origami figures in humans supported the idea that matching hand preferences in the majority of the population evolved due to social learning processes 22.
Importantly, in birds (which were not included in the present analysis), a recent study reported that Psittacine species with stronger left-foot preferences also have larger brains 23. Interestingly, there is some evidence that in Psittaciformes, leftward foot preferences are more common than rightward foot preferences 24,25. This suggests that increased brain size may lead towards a need for coherent lateralization on the side that is the dominant one in most individuals within a species. This implies that no specific evolutionary pressure to either converge to the left or the right side exists.
Several methodological aspects should be considered when interpreting the present results. Importantly, we did not have an equal distribution of animal species over different Mammalian orders, but primate species were clearly over-represented. This was due to data availability but could be problematic since primates tend to have larger brains than most other mammals. Also, there were several marsupial species included in the present study, which are anatomically distinct from placental mammals as they do not have a corpus callosum. While the anterior commissure has a similar function to the corpus callosum in these species and the principal assumptions of the Ringo hypothesis are the same for all Mammalian orders, this anatomical difference may have affected data patterns. Moreover, other factors than brain size may have affected results, for example gyrification.