Mitochondrial Genome Size and D-Loop Variation in Flying, Aquatic, and Terrestrial Mammals: An Exploratory Comparative Analysis

Background: Mitochondrial genome size varies widely among mammals, yet whether this variation is systematically associated with major ecological transitions in locomotor mode remains unclear. Here, we compared complete mitochondrial genome length and the lengths of the D-loop, 12S rRNA, and 16S rRNA among flying, aquatic, and terrestrial mammals. Methods: We used 13,487 complete mitogenome sequences retrieved from NCBI, representing 1674 species across 22 mammalian orders. Non-parametric tests and phylogenetic generalized least squares (PGLS) analyses were used to assess differences and associations. Results: Non-parametric tests showed that all four genomic features differed significantly among ecological groups (Kruskal–Wallis, all p < 0.001). Complete genome length and D-loop length showed concordant patterns (flying > terrestrial > aquatic), whereas rRNA lengths followed distinct patterns, with aquatic mammals possessing the longest 12S and 16S rRNA genes. However, the absolute median differences in rRNA lengths were small (18 bp for 12S rRNA and 12 bp for 16S rRNA), and the 12S rRNA effect size was not negligible (ϵ2 = 0.22), so these differences should not be interpreted as evidence of adaptive divergence. Phylogenetic generalized least squares (PGLS) analyses further showed that ecological group was not significantly associated with any genomic feature after accounting for phylogenetic non-independence (all overall p > 0.10)—a conservative result that, given the very small number of independent ecological transitions, reflects limited statistical power as much as the absence of ecological effects. Under the taxonomic proxy tree, appreciable phylogenetic structure was detected for D-loop length (Pagel’s λ = 0.51) and complete genome size (λ = 0.74), but these values should not be interpreted as strict measures of phylogenetic signal. Conclusions: Because powered flight evolved only once in mammals (Chiroptera) and the fully aquatic lifestyle only twice (Cetacea and Sirenia), these group-level patterns rest on few independent evolutionary transitions and cannot be statistically separated from lineage-specific or ancestral states. We therefore present these findings as an exploratory, hypothesis-generating comparison. Phylogenetically explicit analyses based on a fully resolved species tree, structural characterization of control-region variation, and direct functional assays of replication and transcription efficiency will be required to test the adaptive scenarios proposed here.