News Reader.info

ABSTRACT A longstanding question in evolutionary biology is how natural selection and environmental pressures shape the mitochondrial genomic architectures of organisms. Mitochondria play a

pivotal role in cellular respiration and aerobic metabolism, making their genomes functionally highly constrained. Evaluating selective pressures on mitochondrial genes can provide

functional and ecological insights into the evolution of organisms. Collembola (springtails) are an ancient hexapod group that includes the oldest terrestrial arthropods in the fossil

record, and that are closely associated with soil environments. Of interest is the diversity of habitat stratification preferences (life forms) exhibited by different species within the

group. To understand whether signals of positive selection are linked to the evolution of life forms, we analysed 32 published Collembola mitogenomes in a phylomitogenomic framework. We

found no evidence that signatures of selection are correlated with the evolution of novel life forms, but rather that mutations have accumulated as a function of time. Our results highlight

the importance of nuclear-mitochondrial interactions in the evolution of collembolan life forms and that mitochondrial genomic data should be interpreted with caution, as complex selection

signals may complicate evolutionary inferences. SIMILAR CONTENT BEING VIEWED BY OTHERS POSITIVE SELECTION OVER THE MITOCHONDRIAL GENOME AND ITS ROLE IN THE DIVERSIFICATION OF GENTOO PENGUINS

IN RESPONSE TO ADAPTATION IN ISOLATION Article Open access 08 March 2022 AMERICAN MASTODON MITOCHONDRIAL GENOMES SUGGEST MULTIPLE DISPERSAL EVENTS IN RESPONSE TO PLEISTOCENE CLIMATE

OSCILLATIONS Article Open access 01 September 2020 MITOGENOMIC RESOLUTION OF PHYLOGENETIC CONFLICTS AND ADAPTIVE SIGNATURES IN FELIFORM CARNIVORANS Article 28 May 2025 INTRODUCTION

Understanding the interplay between genomes, morphology, functional ecology, and biogeography in driving organismal evolution has become central to answering fundamental questions in

biology, both from historical (e.g. past speciation) and contemporary (e.g. evolutionary responses to climate change) viewpoints. Making sense of the immense complexity of life has proven a

recalcitrant challenge to biologists. Additional lines of evidence pertaining to taxonomy (the fossil record in conjunction with morphological, behavioural, physiological, and molecular

data) enhance our ability to piece together the diversification of life on Earth1,2,3,4. A central question in molecular evolution revolves around how natural selection influences the

genomic architectures of organisms5,6,7,8,9,10. Environmental pressures drive functional and morphological changes, which are underpinned by genetic variation and, therefore, will manifest

as signatures of selection on the genome that can be used to retrace the evolutionary history of organisms and better predict future adaptive evolutionary trajectories. The mitochondrial

genome comprises 13 protein-coding genes (PCGs), which, together with many nuclear genes, encode the proteins that make up the subunits of the four complexes involved in the oxidative

phosphorylation chain11,12. Due to the crucial role of PCGs in the respiratory chain, these genes are functionally highly constrained5. An evaluation of signals of past selection on these

genes, therefore, provides insights into the adaptive evolution and metabolic requirements of organisms, given that they are responsible for 95% of the energy metabolism and heat production

in cells11,13,14. Mitochondrial genes are some of the most commonly used genetic markers, partly because they provide unique perspectives on population genetic variation and structure due to

maternal inheritance and a general lack of recombination15,16,17, and are frequently used to answer questions related to historical processes and geographical structure13,14,18,19. However,

these inferences often assume that the evolution of this genome is neutral20,21, an idea that has been challenged22,23,24,25,26. It is now widely accepted that the mitogenome is under

selection due to its pivotal role in respiration and energy production, and the evolutionary constraints resulting from the co-evolution of nuclear and mitochondrial genes that are involved

in these processes11,12,15,16,17,25. Additionally, positive selective sweeps, linked to thermal and aerobic respiratory adaptations, are well documented for mitochondrial genes across

various taxa11,13,14,27,28,29,30,31,32. However, the investigation of positive selection on moderately conservative genomic regions, without an assessment of the functional consequence of

amino acid mutations, is not adequate to consider the mitochondrial genome in its entirety to be under positive selection linked to adaptive processes13. Therefore, we identify both signals

of positive selection and the functional and adaptive significance of amino acid mutations on the mitochondrial genomes of ancient soil-dwelling invertebrate taxa. Given the influence of

heterogeneous soil environments on soil biota community assemblages33,34,35, and that positive selection on the mitogenome is linked to thermal and aerobic respiratory adaptations, we sought

to explore the effects of signatures of past selection on the mitogenome of one of the most ancient soil-dwelling invertebrate taxonomic groups, Collembola. Commonly referred to as

springtails, this group of primitive, wingless, and largely terrestrial hexapods is regarded as among the first aquatic arthropods to have colonised land36, with evolutionary roots that

predate the diversification of insects37,38,39,40. Indeed, the earliest known hexapod fossil (from the early Devonian period, ca. 400 Ma) is a collembolan41,42. This group's biology and

ancient evolutionary history provide unique opportunities to understand the environmental and physiological mechanisms governing the evolution of early terrestrial life. Collembolans

currently exploit a myriad of niches across all continents, including Antarctica40,43,44. They typically reside within, or are closely associated with soils, and are considered essential

components of soil ecosystems due to their influence on soil microbial communities, their role in nutrient recycling, and their importance as prey to a wide range of predatory

arthropods45,46,47,48. Of particular interest is the preference of various springtails for different habitats. Springtail species broadly fall into six different ecological life form

categories, each displaying a unique stratification preference. These are atmobiotic (large species that inhabit macrophytes (grasses, bushes, tree trunks and branches)); epiedaphic (species

found in vegetated habitats or the upper litter layer above the soil); hemiedaphic (which includes species that occupy decomposed litter or rotten wood on the soil surface); euedaphic

(‘true’ soil living species that occupy the upper mineral-rich layers of the soil); myrmecophilous (describes species found in ant nests), and hydrophilous (hereafter referred to as aquatic

for simplicity; species that live on the surfaces of, or are closely associated with, freshwater bodies)34,43,47,49,50. Strong correlations exist between the life forms of springtails and

their morphologies, suggesting that life forms are subject to selective forces exerted by their environment43,45,46,47,51,52,53. Additionally, springtails also exhibit highly specific

habitat preferences for different soil types, soil chemistries, habitat types, and stratification within a given habitat47,49,52. It is remarkable that these morphological designs have

probably persisted unchanged for millions of years in many instances. Indeed, springtail fossils dating back as far as the Eocene (ca 40–50 Ma) can be assigned to extant genera and, in some

cases, even extant species54. This extreme morphological conservatism suggests that current morphospecies inventories underestimate true collembolan diversity. Not surprisingly, genetic

evidence has shown that cryptic diversity is pervasive in this group, with current collembolan species richness thought to underestimate the actual number of species by at least an order of

magnitude40,55,56. Therefore, the low number of known springtail species is not an inherent characteristic of the group, but rather the result of under sampling, outdated taxonomy, and a

lack of taxonomic expertise56. While Collembola may be dwarfed by other hexapod groups, particularly insects, it still harbours immense diversity40. Given that Collembola constitutes

ecologically important taxa within soil environments48, which exhibit distinct microhabitat stratification that is driven by feeding strategies, food availability, and soil type and

chemistry47,49,50,52,57, we sought to identify mitochondrial genes that are under positive selection for adaptations to these microhabitats. Additionally, given that metabolic activity and

oxygen consumption differ depending on ecological life form58,59, we investigated whether signals of selection on the mitochondrial genome are associated with the evolution of life form

traits (habitat preferences), in the context of environmental change (oxygen levels/metabolic requirements and vegetation alterations) that emerged during the group's evolutionary

history. If this is the case, we expect substantial signals of selection to have accumulated along phylogenetic branches representing shifts in life forms compared to sister branches where

the original life form was maintained. RESULTS DATED PHYLOGENY AND ANCESTRAL STATE RECONSTRUCTION Our dated mitogenomic phylogeny demonstrates that Collembola first diverged sometime in the

early Devonian (ca. 410 Ma), with most diversification taking place in the Triassic to Cretaceous periods (ca. 250–60 Ma) (Fig. S1). To identify whether a link exists between signals of

mitochondrial DNA selection and collembolan life form (see Dataset S1), we assessed phylogenetic relationships between 32 collembolan and three outgroup taxa (see Table S1). We assigned a

life form trait (i.e. aquatic, epiedaphic, euedaphic, hemiedaphic, and myrmecophilous) to the ecology (i.e. habitat preference) of each springtail species included in this study (see Dataset

S1 for more detail). We reconstructed the ancestral ecological life form state for the collembolan clade using three different methods (parsimony, maximum likelihood, and Bayesian

inference), all of which identified hemiedaphic as the ancestral character (see Fig. 1). Across the collembolan phylogeny, five independent life form shifts occurred from the hemiedaphic to

alternative life forms (see Fig. 1). For the terminal nodes, the most common ecological life form was hemiedaphic (22 taxa), followed by epiedaphic (five taxa), euedaphic (three taxa),

myrmecophilous (one taxon), and aquatic (one taxon). To assess phylogenetic conservatism and homoplasy across the phylogeny, we calculated the retention index (RI)60 for each life form

trait, which was 0.75 for epiedaphic, 0.36 for hemiedaphic, 0.33 for euedaphic, and 0 for both myrmecophilous and aquatic. SELECTION ANALYSIS To assess signals and direction of selection at

different loci (i.e. positive vs purifying selection), we used the codon-based maximum likelihood (CodeML) algorithm to determine the ratio of nonsynonymous to synonymous mutations (ω = 

dN/dS), and the number of nonsynonymous mutations as a proportion of the total number of mutations (dN/(dN + dS))61,62,63,64. Site model selection analyses indicated overall positive

selection across the phylogeny, shown by significant likelihood ratio tests (LRT) for the comparative site models, however, with no nucleotide positions under positive selection (Bayes

Empirical Bayes score (BEB) < 0.9562) (Table S2). The comparison between the numbers of synonymous to nonsynonymous mutations across the terminal nodes of the phylogeny was significantly

different (paired _t_ test: _t_415 = − 9.48, _p_ < 0.05). The number of nonsynonymous mutations as a proportion of the total number of mutations revealed no strong evidence for a link

between life form transitions and the number of nonsynonymous mutations. Departures from the expectation that synonymous mutations are more common than nonsynonymous mutations65 were found

in several taxa (e.g. _Bilobella aurantiaca_, _Neelus murinus_, _Oncopodura yosiiana_, _Podura aquatica_, and _Tullbergia bisetosa_) (see black and white pie charts in Fig. 1 and Table S3

for proportion values > 0.50). However, only the life histories of _Podura aquatica_ and _Tullbergia bisetosa_ deviated from the hemiedaphic ancestral state. The large proportions of

nonsynonymous mutations are likely due to the older ages of these branches and illustrate that the proportion of nonsynonymous mutations increases as a function of time for the taxa with the

ancestral state compared to those that represent life form shifts (i.e. the alternative state) (ancestral state: R2 = 0.52; alternative state: R2 = 0.82; also see Fig. 2). The 95%

confidence intervals of these regressions broadly overlap, and the two models were not significantly different (_p_ > 0.05). This indicates that no significant difference was found in the

proportion of nonsynonymous mutations in relation to branch length for taxa that retained the ancestral state and for those whose ecological life form shifted. The relationships

(represented as coloured pie charts in Fig. 1) between the taxa that represent the life form transitions to alternative life forms and the respective sister taxon/taxa (ancestral state, i.e.

hemiedaphic), again revealed that the longer branches have the greatest number of nonsynonymous mutations in comparison to the sister clades (see coloured pie charts A and D). No

significant difference was found within pie charts B, C, and E, which show that the proportion of mutations for the branches that represent the alternative life form vs. ancestral state (51%

for pie charts B and C, and 50% for pie chart E; paired _t_ test: _t_4 = 0.23, _p_ > 0.05) (Fig. 1). Evidence for purifying (ω < 1) or diversifying selection (ω > 1) was determined

for each of the terminal branches, and a total of 19 taxa had mitogenomes with signatures of purifying selection, while the mitogenomes of 11 taxa showed signs of diversifying selection,

and two taxa had mitogenomes that satisfied the assumptions of neutrality (see Table S3). The proportion of changes per gene and per taxon reveal that _atp8_ exhibits the highest number of

changes for the majority of the taxa (Figs. S2, S3; Table S4). Following _atp8_, all genes from complex I (i.e. _nad1_–_nad6_) had a large proportion of nonsynonymous mutations. Of the 13

PCGs, eight genes were identified as experiencing purifying selection, while five genes were identified as experiencing diversifying selection, with _atp8_ being under particularly strong

diversifying selection (ω > 3.0) (Table S4). At a complex level, complex V (ATPase) revealed the highest percentage of nonsynonymous mutations, followed by complex I (NDs), complex IV

(COXs), and complex III (_cytb_) (Fig. S3), which follows the estimated mutation rate of these complexes (i.e. ATPase > ND > COX > CYTB)66. The average value of ω for all genes

across all terminal nodes was 0.96, which indicates that, overall, Collembola mitogenomes are experiencing slight purifying selection or no selection (i.e. selective neutrality). RADICAL

CHANGES IN PHYSICOCHEMICAL PROPERTIES OF AMINO ACIDS The impact of positive selection on the physical and chemical properties of 31 amino acids was evaluated for each of the 13 PCGs by

detecting observed changes inferred by the phylogenetic tree under neutral conditions. Both positive and negative significant z-scores on the physicochemical properties of amino acids were

found across the tree. Most physicochemical properties indicated that the radical changes were under purifying selection, with occasional radical changes under positive selection (Fig. S4).

We detected two physicochemical properties—equilibrium constant (ionisation of COOH) and solvent accessible reduction ratio—with significant radical changes (z > 3.09; _p_ < 0.001),

indicating positive/diversifying selection for all 13 PCGs, while a third physicochemical property, buriedness, is under positive/diversifying selection for 10 PCGs (except _cytb_, _nad3_,

and _nad4l_), with _atp8_ having the highest z-score. DISCUSSION Using a comprehensive mitogenome dataset for springtails (32 species from all four orders), we determined the ancestral

ecological life form state for Collembola. We then used this dataset to assess whether a correlation exists between signatures of mitogenome selection and the evolution of novel life forms.

We show that the ancestral ecological life form state of Collembola is hemiedaphic, with epiedaphic life forms first emerging in the early Permian, roughly 290 Ma. Even though several

previous studies have identified links between Collembolan mitogenomes and ecology, morphology, and physiology50,53,67,68,69,70,71, we found no strong evidence for a link between signatures

of mitochondrial genome selection and life form evolution in Collembola. Our mitogenomic phylogeny is broadly consistent with those presented in earlier studies37,38,39,72,73,74,75. The

first phylogenetic split within Collembola occurred during the early Devonian (ca. 410 Ma), with most diversification taking place in the Triassic to Cretaceous periods (ca. 250–60 Ma). It

is not surprising, therefore, that Collembola fossils preserved in amber and dating back to the Eocene (50-40 Ma) can be assigned to extant taxa with high certainty54, demonstrating the

ultraconservative nature of life forms and basic body plans of the group. Ancestral state reconstructions showed the ancestral ecological life form of collembolans to be hemiedaphic. Our

finding contrasts with the assumption of a euedaphic origin of springtails proposed by D’Haese36,76. Additionally, we found that the epiedaphic life form evolved at least twice, first during

the origin of Symphypleona (Fig. S1), approximately 290 Ma, and then again in the mid-Cretaceous with the emergence of the evolutionary lineage represented by _Salina celebensis_, roughly

115 Ma. During the Permian period, roughly 300 Ma, seed-producing gymnosperms such as gnetophytes, Pinaceae, and cupressophytes diverged from cycads and conifers77. Symphypleona show a

habitat preference for Mediterranean-type shrublands, pine plantations, mixed forests, bushes, grasslands, and leaf litter (see Supplementary Information Dataset S1). This suggests that the

emergence of major plant lineages, such as gymnosperms, created new ecological niches for the epiedaphic Symphypleona to exploit and diversify. The extant _Bourletiella arvalis_ and

_Sminthurus viridis_, however, show a habitat preference for grasslands, although these habitats would have only become available much later in the Cenozoic78,79,80. It is, therefore,

reasonable to assume that the evolutionary lineage represented by these species would have occupied gymnosperm ecological analogs until their current niches became available. Unsurprisingly,

the morphological features of these taxa are characteristic of the epiedaphic life form47,49,81. Additionally, our analyses illustrate that the euedaphic life form evolved at least twice,

first when _Tullbergia bisetosa_ evolved, roughly 287 Ma, and secondly during the origin of _Onychiurus orientalis_ and _Orthonychiurus folsomi_ (approximately 192 Ma). A shift from

hemiedaphic to the myrmecophilous and aquatic life forms was also identified in the phylogeny. The origin of _Cyphoderus albinus_ and the myrmecophilous life form occurred approximately 170 

Ma, coinciding with the evolution of ants (Formicidae) around 168 Ma82,83. The aquatic life form first emerged about 225 Ma with the divergence of the lineage represented by _Podura

aquatica_. The hypothesis of a semi-aquatic or aquatic life form as the ancestral state of Collembola was formulated and tested by D’Haese36,76, who found that the evolution of terrestrial

to aquatic habitats, in a physiological context, is indeed more plausible and that an aquatic life form can be viewed as a “rare secondary acquisition”36,43,76. We deem the five life forms

we have identified to be somewhat homoplastic (see Yu et al.53) best illustrated by the epiedaphic and euedaphic forms that independently originated twice within the group. Additionally, the

myrmecophilous and aquatic forms are homoplastic or autapomorphic, as these traits are unique to single taxa. There is also a great deal of ecological conservatism for the hemiedaphic trait

where most extant taxa have retained the ancestral state. While there appeared to be a general trend for taxa on longer phylogenetic branches to have a greater overall proportion of

nonsynonymous mutations, we found no consistent relationship between life form shifts and the proportion of nonsynonymous mutations on these branches. Furthermore, even when comparing

individual genes, no correlations were found between the proportion of nonsynonymous changes and the evolution of new life forms (Fig. S2). This may seem surprising given the pivotal role

that mitochondria play in aerobic metabolism, and given that different life forms are constrained in different ways metabolically. It should be noted, however, that for a group as ancient as

Collembola, the many adaptive nuclear changes that produced the conspicuous ecological changes upon which life forms are classified may have obscured mitochondrial signatures of selection

related to the evolution of life forms. While our expectations pertaining to selection were premised on non-overlapping habitat stratification preferences, it should also be noted that

interactions between collembolans and their environments are complex, and apparently divergent life forms may impose similar functional constraints on the organism, thereby confounding

signatures of selection84. Even though mitochondrial protein coding genes are typically believed to have faster mutation rates than nuclear genes16, most of these mutations are probably

synonymous85,86, and thus functionally inconsequential (but see Lawrie et al.87). Our results support this idea, with more taxa and more genes under purifying selection than under

diversifying selection (Tables S3 and S4). However, selection on the mitochondrial genome appears to be largely neutral due to the random fixation of selectively neutral mutations21. Given

the crucial role of mitochondria in aerobic metabolism, functional requirements likely constrain the molecular evolution of mitogenomes, and this is perhaps why many studies have identified

purifying selection88,89,90. It makes sense that strong functional constraints would mean that any deleterious mutation that arises would be rapidly removed from the population via purifying

selection. Additionally, genetic drift plays a major role in structuring genetic diversity, and its effects on mitochondrial genomes are stronger than on nuclear genomes due to differences

in effective population sizes between them and the uniparental inheritance and lack of recombination within the mitochondrial genome16,17,21. When radical amino acid changes are favoured by

selection, local directional shifts in the structure and function of proteins evolve. The physicochemical properties under positive selection for all PCGs were equilibrium constant

(ionisation of COOH) and solvent accessible reduction ratio, while buriedness was under positive selection for 10 of the 13 PCGs. Positive selection for equilibrium constant is suggestive of

adaptations to increased oxygen levels and metabolic requirements66, while positive selection for solvent accessible reduction ratio may lead to bulkier proteins, which increases the space

for active site formation66. Buriedness increases protein stability, and thus optimal protein function, in the context of hydrophobicity, whereby some residues are buried on the interior of

the globular protein91,92. These results illustrate the importance of these properties in the context of environmental conditions (oxygen levels and metabolic requirements). The

mitochondrial genome has long been considered to be selectively neutral22,23,24,25, but a plethora of studies have illustrated that it may be under strong positive selection as a result of

thermal and aerobic respiratory adaptations11,13,14,27,28,29,30,31,32. Here, we investigated whether a link exists between mitochondrial genome selection and the evolution of ecological life

forms in Collembola. The expected outcome of such a link would be evidence for strong signals of selection at branches representing the evolution of new life forms, even if mitochondrial

genes themselves are not the primary drivers of ecophysiological adaptation. Our selection results reveal that, overall, positive selection is present across the collembolan phylogeny, best

illustrated by the presence of positive selection on three amino acid physicochemical properties which likely play instrumental roles in adaptive responses to environmental conditions

(oxygen levels and metabolic requirements). However, the only major trend identified here was that the proportion of nonsynonymous mutations increased with time, with longer branches (i.e.

older taxa) having a greater proportion of nonsynonymous mutations than shorter branches. In addition to investigating selection associated with life form evolution, we assessed whether

there was a link between signatures of selection across a latitudinal gradient from where each species can be found. This was done in two ways; first by selecting species across all

latitudes but where their ranges overlapped, and secondly by selecting species that had no overlapping distributions. As was the case for life form shifts, we found no clear correlation

between the number of nonsynonymous mutations and latitudinal gradients for either analysis. The lack of a correlation between life form shifts and mitogenome selection may hint at nuclear

selection, and suggests that nuclear-mitochondrial interactions are of greater importance in the evolution of life form shifts than selection pressures on the mitochondrial genome alone. Our

findings raise some precautionary flags in that selection, where present, is more complex than generally assumed, given the complex interactions of the mitochondrial and nuclear genomes.

Importantly, evidence for strong selection on the mitochondrial genome cannot simply be extrapolated to reflect adaptation to changing environmental conditions. METHODS PHYLOMITOGENOMIC

ANALYSIS Phylomitogenomic relationships between Collembola taxa based on available data on GenBank (May 2020) were reconstructed using the protein-coding genes (PCGs) of 32 partial or

complete Collembola mitogenomes, as well as three outgroup taxa (see Supplementary Information Table S1). Each PCG was aligned using the MUSCLE algorithm in MEGA X93, followed by

concatenating the alignments into a single dataset. Substitution saturation was assessed on each codon position using DAMBE v7.3.594, which revealed high substitution saturation, likely due

to highly divergent taxa95,96,97. As a result, we took a conservative approach and used the amino acid sequences of each PCG for our phylogenetic reconstructions. Each PCG was independently

translated into amino acid sequences using the invertebrate mitochondrial genetic code, and were then aligned in MEGA X using the MUSCLE algorithm, followed by manual inspection. Bayesian

phylogenetic inferences were executed in BEAST v2.6.398 by linking site and clock models as well as trees. This was done because of the largely uniparental inheritance of mitochondrial

genomes and a lack of recombination15,16,17,27. A discrete trait was added to define the ecological life form of each species (see Supplementary Information Dataset S1 for more detail).

Given the role of mitochondria in different environments, and rather than linking life forms to morphological characteristics, we linked the life forms (i.e. aquatic, myrmecophilous,

euedaphic, hemiedaphic, and epiedaphic) to the ecology (i.e. habitat preferences) of each springtail species included in this study however, with the exception of three taxa; _Onychiurus

orientalis_, _Orthonychiurus folsomi_, and T_ullbergia bisetosa_, which were assigned based on morphology. These species, although ecologically hemiedaphic47,49,50 (see Dataset S1 for

ecological life forms), differ from other hemiedaphic taxa by having all three major euedaphic morphological characteristics (no eyes, no pigment, and no furca), and are thus categorised as

“euedaphomorphic” sensu lato99. As such, these taxa were assigned to the euedaphic life form which aligns to a peculiar ecology within litter (i.e. lower dispersal ability and associated

traits) that differs from the other hemiedaphic taxa. The mitochondrial REV + Г model was selected, with an uncorrelated relaxed lognormal clock and a birth–death tree prior. Monophyly was

constrained for several clades based on a consensus tree constructed using the NCBI taxonomy database in PhyloT v2 (https://phylot.biobyte.de/), and visualised in iTOL100. These were (a) all

taxa (to calibrate the root of the tree, as described below), (b) the Hexapoda clade (all taxa excluding the distantly related _Daphnia pulex_) to calibrate the split between Collembola and

other basal hexapod taxa (_Japyx solifugus_ and _Trigoniophthalmus alternatus_) and Collembola as described below, and (c) a clade that excluded all non-Collembola taxa. Monophyly was not

constrained for each collembolan order (Poduromorpha, Entomobryomorpha, Neelipleona, and Symphypleona) since monophyly is not always supported for these groups101,102,103. Considering that

some collembolan fossils have been assigned to contemporary species despite being more than 20 Ma old, dating recent nodes using fossil data can be challenging for these taxa. We therefore

used secondary calibration points from Leo et al.104 and included calibration points that matched the two deepest fossil-calibrated nodes. To ensure that our calibration points matched the

95% highest probability density (HPD) confidence intervals of Leo et al.104 for the same nodes, we followed their protocol in setting the priors for these nodes. Accordingly, we set the

prior at the root of the tree to have a normal distribution and a mean of 510 Ma with a standard deviation of 7 Ma. The second prior was placed at the split between the higher insects

(_Japyx solifugus_ and _Trigoniophthalmus alternatus_) and Collembola, with a normal distribution and a mean of 485 Ma with a standard deviation of 6 Ma (for the rationale behind this prior

assignment, see Leo et al.104). A Bayesian consensus phylogenetic tree was calculated using BEAST submitted through the CIPRES Science Gateway105 from three independent chains at different

starting seeds for a chain length of 280 million generations, a thinning interval of 10,000, and 25% burn-in to ensure that stationarity was reached. Convergence was assessed using Tracer

v1.7106, trees were summarised using TreeAnnotator v2.6.298 (20% burn-in), and the resulting topology was visualised in FigTree v1.4.3 (http://tree.bio.ed.ac.uk/software/figtree/). ANCESTRAL

STATE RECONSTRUCTION A quantitative phyletic approach was used to assess the evolutionary relationships between taxa based on this biological information. Ancestral state reconstructions

were performed using both parsimony and maximum likelihood methods in Mesquite v3.61107; the discrete trait of each taxon (from Supplementary Information Dataset S1) was provided as an input

for the analyses. The likelihood ratio test of the maximum likelihood method was used for the Markov k-state one-parameter (Mk1) and the asymmetrical Markov k-state 2 parameter (AsymmMk)

models107,108, while the rate of character evolution was estimated under the Mk1 model. The Bayesian consensus phylogenetic tree obtained from the 28,000 trees (after the burn-in) sampled

from BEAST was used for the ancestral state reconstruction. The retention index (RI)60 was evaluated for each life form trait to assess phylogenetic conservatism and homoplasy across the

phylogeny. High RI values represent little to no homoplasy (typically above 0.85, see Yu et al.53), while low RI values indicate homoplasy and low phylogenetic signal in a given dataset. To

account for uncertainties with phylogenetic mapping and the occurrence of topographical incongruence across sampled trees, a Bayesian MCMC method implemented in BayesTraits v3.0.2109, was

used for ancestral state reconstructions. The _AddTag _and_ AddMRCA_ options were implemented to find the ancestral state associated with the node of collembolan clade. These options were

used to overcome uncertainty when estimating ancestral states109. A reversible-jump (RJ) hyper-prior was implemented to reduce uncertainty and arbitrariness when selecting priors109. The

hyper-prior draws up values from a uniform distribution between 0 and 30 which is used to seed the exponential distribution for the MCMC analyses. Three independent MCMC chains were run for

30 million iterations with a 25% burn-in. These analyses were repeated with a gamma reversible-jump hyper-prior (mean: 0, 30; variance 0, 30), but no appreciable difference was identified

between the posterior distributions between these analyses when visualised in Tracer. MCMC convergence was assessed in Tracer by ensuring that the effective sample size (ESS) ≥ 200 for all

estimated parameters. The proportion of each state was calculated from the mean values from approximately 20,000 trees randomly chosen by BayesTraits from the 28,000 trees sampled from

BEAST. The states that were reconstructed to be equivocal (probability of 0.2 for each state) were discarded and the probability that the collembolan clade was in a given state was

determined. The average of the estimated probability for the three independent runs was calculated, and any character state with a proportion of > 0.7 was considered to be supported110.

SELECTION ANALYSIS The ratio of nonsynonymous, (dN; i.e. amino acid-altering) to synonymous (dS; i.e. silent) substitution rates, ω = dN/dS, is used to understand the evolutionary dynamics

of genes61,62,63,64. We computed ω using the codon-based maximum likelihood (CodeML) algorithm in EasyCodeML v1.2162,111. To evaluate the influence of selection on the PCG sequence

alignments, we implemented the preset (nested models) running mode with the site model within the framework of the topology of the phylogeny retrieved by the phylogenetic analyses above112.

The codon substitution models described by Geo et al.62, and the significance of their respective dN/dS ratios, were evaluated using the likelihood ratio test (LRT). To identify whether a

link between signatures of mitochondrial selection and a shift in ecological life form exists, we determined the number of nonsynonymous mutations per gene as a proportion of the total

mutations (i.e. dN/(dN + dS)) for all terminal branches, and then averaged these proportions per branch. Moreover, we compared the proportion of nonsynonymous mutations for each life form

shift (aquatic, myrmecophilous, euedaphic, and epiedaphic) vs. the respective sister branches/taxa that represented the ancestral state (hemiedaphic) to ascertain whether a link between the

number nonsynonymous mutations to the ecological life form transition exists. Where multiple branches were present (clades), we accounted for the number of taxa by averaging the proportion

of mutations for each terminal branch, and then followed the path from the terminal branch back until reaching the node where life form divergence (i.e. nearest common ancestor) took place.

Statistical analyses were performed using a paired _t_ test to compare the proportion of synonymous to nonsynonymous mutations across the phylogeny as well as where each branch is paired

with its sister branch that represents a life form shift. Plots of regression analyses were performed using the R package ggplot2113 to assess the correlation between the proportions of

nonsynonymous changes vs. branch length for the taxa that retained the ancestral state and for those who represent a life form shift (i.e. the alternative state). Univariate general linear

models of the regressions were performed in SPSS 27 (Inc., Chicago, IL, USA). Furthermore, and following the same method as described above, we assessed whether there was a link between

selection and species that originated from different latitudinal gradients. This was done twice, first by identifying species across latitudinal gradients where their ranges overlapped, and

secondly by selecting species that had no overlapping distributions. A phylogenetic tree was reconstructed for both analyses, and selection analyses were conducted, and the number of

nonsynonymous mutations were determined for each terminal branch/species. Additionally, we sought to determine the evolutionary dynamics in the context of purifying or diversifying selection

for each terminal branch based on the count of nonsynonymous and synonymous mutations (i.e. ω) for each gene and complex. To represent these results to a gene or complex level (in the form

of a percentage), the number of changes per gene/complex were standardised based on gene size (i.e. total number of nonsynonymous changes/(gene size × total number of branches) × 100).

RADICAL CHANGES IN PHYSICOCHEMICAL PROPERTIES OF AMINO ACIDS The impact of positive selection on physical and chemical properties of 31 amino acids was evaluated for each of the 13 PCGs

using TreeSAAP v3.2 (Selection on Amino Acid Properties)114 by detecting observed changes inferred by the phylogenetic tree under neutral conditions. The impact of 20 physicochemical

properties (excluding 11 amino acids with < 85% accuracy following McClellan and Ellison115) were evaluated. Under the assumption of neutrality, a significant positive z-score as

calculated by TreeSAAP (z-score > 3.09, _p_ < 0.001) for a physicochemical property is indicative of the prevalence of nonsynonymous substitutions, while a significant negative z-score

(z-score < − 3.09, _p_ < 0.001) indicates the prevalence of synonymous substitutions. A sliding window of 15 codons was utilized, based on a categorical scale of 1 (most conservative)

to 8 (most radical). Only the most radical changes (category scores of 7 and 8, _p_ ≤ 0.001) were considered, and for the physicochemical properties that were significant for both category

7 and 8, only the z-scores that were significant for category 8 were included. DATA AVAILABILITY The authors confirm that all relevant data used for this manuscript are available online from

the GenBank NCBI online repository (https://www.ncbi.nlm.nih.gov/genbank/). All sequence accession numbers are list in the Supplementary Information of this manuscript (Table S1).

REFERENCES * Zachos, F. E. (New) Species concepts, species delimitation and the inherent limitations of taxonomy. _J. Genet._ 97, 811–815. https://doi.org/10.1007/s12041-018-0965-1 (2018).

Article  PubMed  Google Scholar  * Zachos, F. E. _et al._ Species inflation and taxonomic artefacts—A critical comment on recent trends in mammalian classification. _Mamm. Biol._ 78, 1–6.

https://doi.org/10.1016/j.mambio.2012.07.083 (2013). Article  Google Scholar  * De Queiroz, K. Species concepts and species delimitation. _Syst. Bot._ 56, 879–886.

https://doi.org/10.1080/10635150701701083 (2007). Article  Google Scholar  * Morard, R. _et al._ Nomenclature for the nameless: A proposal for an integrative molecular taxonomy of cryptic

diversity exemplified by planktonic Foraminifera. _Syst. Biol._ 65, 925–940. https://doi.org/10.1093/sysbio/syw031 (2016). Article  PubMed  Google Scholar  * Shtolz, N. & Mishmar, D. The

mitochondrial genome—On selective constraints and signatures at the organism, cell, and single mitochondrion levels. _Front. Ecol. Evol._ 7, 342. https://doi.org/10.3389/fevo.2019.00342

(2019). Article  Google Scholar  * Pease, J. B., Haak, D. C., Hahn, M. W. & Moyle, L. C. Phylogenomics reveals three sources of adaptive variation during a rapid radiation. _PLoS Biol._

14, e1002379. https://doi.org/10.1371/journal.pbio.1002379 (2016). Article  CAS  PubMed  PubMed Central  Google Scholar  * Rosenblum, E. B., Römpler, H., Schöneberg, T. & Hoekstra, H. E.

Molecular and functional basis of phenotypic convergence in white lizards at White Sands. _Proc. Natl. Acad. Sci. U.S.A._ 107, 2113–2117. https://doi.org/10.1073/pnas.0911042107 (2010).

Article  ADS  PubMed  Google Scholar  * Rosenblum, E. B., Parent, C. E. & Brandt, E. E. The molecular basis of phenotypic convergence. _Annu. Ecol. Evol. Syst._ 45, 203–226.

https://doi.org/10.1146/annurev-ecolsys-120213-091851 (2014). Article  Google Scholar  * Brawand, D. _et al._ The genomic substrate for adaptive radiation in African cichlid fish. _Nature_

513, 375–381. https://doi.org/10.1038/nature13726 (2014). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Dunn, C. W. & Munro, C. Comparative genomics and the diversity of

life. _Zool. Scr._ 45, 5–13. https://doi.org/10.1111/zsc.12211 (2016). Article  Google Scholar  * Finch, T. M., Zhao, N., Korkin, D., Frederick, K. H. & Eggert, L. S. Evidence of

positive selection in mitochondrial complexes I and V of the African elephant. _PLoS One_ 9, e92587. https://doi.org/10.1371/journal.pone.0092587 (2014). Article  ADS  CAS  PubMed  PubMed

Central  Google Scholar  * van den Heuvel, L. P. & Smeitink, J. A. M. Nuclear DNA and oxidative phosphorylation. In _Oxidative Phosphorylation in Health and Disease_ (eds Smeitink, J. A.

M. _et al._) 117–129 (Springer, 2004). https://doi.org/10.1007/0-387-26992-4_7. Chapter  Google Scholar  * da Fonseca, R. R., Johnson, W. E., O’Brien, S. J., Ramos, M. J. & Antunes, A.

The adaptive evolution of the mammalian mitochondrial genome. _BMC Genomics_ 9, 1–22. https://doi.org/10.1186/1471-2164-9-119 (2008). Article  CAS  Google Scholar  * Teacher, A. G. F.,

André, C., Merilä, J. & Wheat, C. W. Whole mitochondrial genome scan for population structure and selection in the Atlantic herring. _BMC Evol. Biol._ 12, 1–14.

https://doi.org/10.1186/1471-2148-12-248 (2012). Article  Google Scholar  * Pesole, G., Gissi, C., De Chirico, A. & Saccone, C. Nucleotide substitution rate of mammalian mitochondrial

genomes. _J. Mol. Evol._ 48, 427–434. https://doi.org/10.1007/pl00006487 (1999). Article  ADS  CAS  PubMed  Google Scholar  * Ballard, J. W. O. & Whitlock, M. C. The incomplete natural

history of mitochondria. _Mol. Ecol._ 13, 729–744. https://doi.org/10.1046/j.1365-294x.2003.02063.x (2004). Article  PubMed  Google Scholar  * Galtier, N., Nabholz, B., Glemin, S. &

Hurst, G. D. Mitochondrial DNA as a marker of molecular diversity: A reappraisal. _Mol. Ecol._ 18, 4541–4550. https://doi.org/10.1111/j.1365-294x.2009.04380.x (2009). Article  CAS  PubMed 

Google Scholar  * Moritz, C. Applications of mitochondrial DNA analysis in conservation: A critical review. _Mol. Ecol._ 3, 401–411. https://doi.org/10.1111/j.1365-294x.1994.tb00080.x

(1994). Article  CAS  Google Scholar  * Avise, J. C. & Ellis, D. Mitochondrial DNA and the evolutionary genetics of higher animals. _Philos. Trans. R. Soc. B Biol. Sci._ 312, 325–342.

https://doi.org/10.1098/rstb.1986.0011 (1986). Article  ADS  CAS  Google Scholar  * Kimura, M. _The Neutral Theory of Molecular Evolution_ (Cambridge University Press, 1983). Book  Google

Scholar  * Kimura, M. The neutral theory of molecular evolution: A review of recent evidence. _J. Jpn. Genet._ 66, 367–386. https://doi.org/10.1266/jjg.66.367 (1991). Article  CAS  Google

Scholar  * Ballard, J. W. O. & Kreitman, M. Is mitochondrial DNA a strictly neutral marker?. _Trends Ecol. Evol._ 10, 485–488. https://doi.org/10.1016/s0169-5347(00)89195-8 (1995).

Article  CAS  PubMed  Google Scholar  * Rand, D. M. The units of selection on mitchondrial DNA. _Annu. Rev. Ecol. Syst._ 32, 415–448. https://doi.org/10.1146/annurev.ecolsys.32.081501.114109

(2001). Article  Google Scholar  * Ballard, J. W. O. & Kreitman, M. Unraveling selection in the mitochondrial genome of _Drosophila_. _Genetics_ 138, 757–772.

https://doi.org/10.1093/genetics/138.3.757 (1994). Article  CAS  PubMed  PubMed Central  Google Scholar  * Ballard, J. W. O. & Rand, D. M. The population biology of mitochondrial DNA and

its phylogenetic implications. _Annu. Ecol. Evol. Syst._ 36, 621–642. https://doi.org/10.1146/annurev.ecolsys.36.091704.175513 (2005). Article  Google Scholar  * Teske, P. R. _et al._

Mitochondrial DNA is unsuitable to test for isolation by distance. _Sci. Rep._ 8, 1–9. https://doi.org/10.1038/s41598-018-25138-9 (2018). Article  CAS  Google Scholar  * Shen, Y.-Y. _et al._

Adaptive evolution of energy metabolism genes and the origin of flight in bats. _Proc. Natl. Acad. Sci._ 107, 8666–8671. https://doi.org/10.1073/pnas.0912613107 (2010). Article  ADS  PubMed

  PubMed Central  Google Scholar  * Yu, L., Wang, X., Ting, N. & Zhang, Y. Mitogenomic analysis of Chinese snub-nosed monkeys: Evidence of positive selection in NADH dehydrogenase genes

in high-altitude adaptation. _Mitochondrion_ 11, 497–503. https://doi.org/10.1016/j.mito.2011.01.004 (2011). Article  CAS  PubMed  Google Scholar  * Silva, G., Lima, F. P., Martel, P. &

Castilho, R. Thermal adaptation and clinal mitochondrial DNA variation of European anchovy. _Proc. R. Soc. B_ 281, 20141093. https://doi.org/10.1098/rspb.2014.1093 (2014). Article  CAS 

PubMed  PubMed Central  Google Scholar  * Consuegra, S., John, E., Verspoor, E. & de Leaniz, C. G. Patterns of natural selection acting on the mitochondrial genome of a locally adapted

fish species. _Genet. Sel. Evol._ 47, 1–10. https://doi.org/10.1186/s12711-015-0138-0 (2015). Article  CAS  Google Scholar  * Li, X. _et al._ Positive selection drove the adaptation of

mitochondrial genes to the demands of flight and high-altitude environments in grasshoppers. _Front. Genet._ 9, 605. https://doi.org/10.3389/fgene.2018.00605 (2018). Article  CAS  PubMed 

PubMed Central  Google Scholar  * Lajbner, Z., Pnini, R., Camus, M. F., Miller, J. & Dowling, D. K. Experimental evidence that thermal selection shapes mitochondrial genome evolution.

_Sci. Rep._ 8, 1–12. https://doi.org/10.1038/s41598-018-27805-3 (2018). Article  CAS  Google Scholar  * Briones, M. J. I. The serendipitous value of soil fauna in ecosystem functioning: The

unexplained explained. _Front. Environ. Sci._ 6, 149. https://doi.org/10.3389/fenvs.2018.00149 (2018). Article  Google Scholar  * Orgiazzi, A. _et al._ Geographical and temporal

distribution. In _Global Soil Biodiversity Atlas_ (eds Orgiazzi, A. _et al._) 66–91 (European Commission, Publications Office of the European Union, 2016). https://doi.org/10.2788/2613.

Chapter  Google Scholar  * Bardgett, R. D., Yeates, G. W. & Anderson, J. M. Patterns and determinants of soil biological diversity. In _Biological Diversity and Functions in Soils_ (eds

Bardgett, R. D. _et al._) 100–118 (Cambridge University Press, 2005). https://doi.org/10.1017/cbo9780511541926.007. Chapter  Google Scholar  * D’Haese, C. A. Were the first springtails

semi-aquatic? A phylogenetic approach by means of 28S rDNA and optimization alignment. _Proc. R. Soc. Lond. B_ 269, 1143–1151. https://doi.org/10.1098/rspb.2002.1981 (2002). Article  CAS 

Google Scholar  * Kristensen, N. P. Phylogeny of insect orders. _Annu. Rev. Entomol._ 26, 135–157. https://doi.org/10.1146/annurev.en.26.010181.001031 (1981). Article  Google Scholar  *

Carapelli, A., Nardi, F., Dallai, R. & Frati, F. A review of molecular data for the phylogeny of basal hexapods. _Pediobiologia_ 50, 191–204. https://doi.org/10.1016/j.pedobi.2006.01.001

(2006). Article  CAS  Google Scholar  * Nardi, F. _et al._ Hexapod origins: Monophyletic or paraphyletic?. _Science_ 299, 1887–1889. https://doi.org/10.1126/science.1078607 (2003). Article

  ADS  CAS  PubMed  Google Scholar  * Cicconardi, F., Fanciulli, P. P. & Emerson, B. C. Collembola, the biological species concept and the underestimation of global species richness.

_Mol. Ecol._ 22, 5382–5396. https://doi.org/10.1111/mec.12472 (2013). Article  CAS  PubMed  Google Scholar  * Whalley, P. & Jarzembowski, E. A. A new assessment of _Rhyniella_, the

earliest known insect, from the Devonian of Rhynie, Scotland. _Nature_ 291, 317. https://doi.org/10.1038/291317a0 (1981). Article  ADS  Google Scholar  * Hirst, S. & Maulik, S. On some

arthropod remains from the Rhynie Chert (old red sandstone). _Geol. Mag._ 63, 69–71. https://doi.org/10.1017/s0016756800083692 (1926). Article  ADS  Google Scholar  * Deharveng, L., D’Haese,

C. A. & Bedos, A. Global diversity of springtails (Collembola; Hexapoda) in freshwater. _Hydrobiologia_ 595, 329–338. https://doi.org/10.1007/s10750-007-9116-z (2008). Article  Google

Scholar  * Hopkin, S. P. _Biology of the Springtails (Insecta: Collembola)_ (Oxford University Press, 1997). Google Scholar  * Ponge, J., Dubs, F., Gillet, S., Sousa, J. P. & Lavelle, P.

Decreased biodiversity in soil springtail communities: The importance of dispersal and landuse history in heterogeneous landscapes. _Soil Biol. Biochem._ 38, 1158–1161.

https://doi.org/10.1016/j.soilbio.2005.09.004 (2006). Article  CAS  Google Scholar  * Auclerc, A., Ponge, J. F., Barot, S. & Dubs, F. Experimental assessment of habitat preference and

dispersal ability of soil springtails. _Soil Biol. Biochem._ 41, 1596–1604. https://doi.org/10.1016/j.soilbio.2009.04.017 (2009). Article  CAS  Google Scholar  * Salmon, S. _et al._ Linking

species, traits and habitat characteristics of Collembola at European scale. _Soil Biol. Biochem._ 75, 73–85. https://doi.org/10.1016/j.soilbio.2014.04.002 (2014). Article  CAS  Google

Scholar  * Rusek, J. Biodiversity of Collembola and their functional role in the ecosystem. _Biodivers. Conserv._ 7, 1207–1219. https://doi.org/10.1023/a:1008887817883 (1998). Article 

Google Scholar  * Potapov, A. A., Semenina, E. E., Korotkevich, A. Y., Kuznetsova, N. A. & Tiunov, A. V. Connecting taxonomy and ecology: Trophic niches of collembolans as related to

taxonomic identity and life forms. _Soil Biol. Biochem._ 101, 20–31. https://doi.org/10.1016/j.soilbio.2016.07.002 (2016). Article  CAS  Google Scholar  * Malcicka, M., Berg, M. P. &

Ellers, J. Ecomorphological adaptations in Collembola in relation to feeding strategies and microhabitat. _Eur. J. Soil Biol._ 78, 82–91. https://doi.org/10.1016/j.ejsobi.2016.12.004 (2017).

Article  Google Scholar  * Parisi, V., Menta, C., Gardi, C., Jacomini, C. & Mozzanica, E. Microarthropod communities as a tool to assess soil quality and biodiversity: A new approach in

Italy. _Agric. Ecosyst. Environ._ 105, 323–333. https://doi.org/10.1016/j.agee.2004.02.002 (2005). Article  Google Scholar  * Martins da Silva, P. _et al._ Traits of collembolan life-form

indicate land use types and soil properties across an European transect. _Appl. Soil Ecol._ 97, 69–77. https://doi.org/10.1016/j.apsoil.2015.07.018 (2016). Article  Google Scholar  * Yu, D.

_et al._ Molecular phylogeny and trait evolution in an ancient terrestrial arthropod lineage: Systematic revision and implications for ecological divergence (Collembola, Tomocerinae). _Mol.

Phylogenet. Evol._ 154, 106995. https://doi.org/10.1016/j.ympev.2020.106995 (2020). Article  PubMed  Google Scholar  * Robin, N., D’Haese, C. & Barden, P. Fossil amber reveals

springtails’ longstanding dispersal by social insects. _BMC Evol. Biol._ 19, 1–12. https://doi.org/10.1186/s12862-019-1529-6 (2019). Article  CAS  Google Scholar  * Cicconardi, F., Nardi,

F., Emerson, B. C., Frati, F. & Fanciulli, P. P. Deep phylogeographic divisions and long-term persistence of forest invertebrates (Hexapoda: Collembola) in the North-Western

Mediterranean basin. _Mol. Ecol._ 19, 386–400. https://doi.org/10.1111/j.1365-294x.2009.04457.x (2010). Article  CAS  PubMed  Google Scholar  * Deharveng, L. Recent advances in Collembola

systematics. _Pediobiologia_ 48, 415–433. https://doi.org/10.1016/j.pedobi.2004.08.001 (2004). Article  Google Scholar  * Berg, M. P., Kniese, J., Bedaux, J. & Verhoef, H. A. Dynamics

and stratification of functional groups of micro- and mesoarthropods in the organic layer of a Scots pine forest. _Biol. Fertil. Soils_ 26, 268–284. https://doi.org/10.1007/s003740050378

(1998). Article  Google Scholar  * Petersen, H. Population dynamic and metabolic characterization of Collembola species in a beech forest ecosystem. In _Soil Biology as Related to Land Use

Practices_ (ed. Dindal, D. L.) 806–833 (Office of Pesticide and Toxic Substance, 1980). Google Scholar  * Coulibaly, S. F. M. _et al._ Functional assemblages of Collembola determine soil

microbial communities and associated functions. _Front. Environ. Sci._ 7, 52. https://doi.org/10.3389/fenvs.2019.00052 (2019). Article  Google Scholar  * Farris, S. The retention index and

the rescaled consistency index. _Cladistics_ 5, 417–419. https://doi.org/10.1111/j.1096-0031.1989.tb00573.x (1989). Article  PubMed  Google Scholar  * Yang, Z. & Bielawski, J. P.

Statistical methods for detecting molecular adaptation. _Trends Ecol. Evol._ 15, 496–503. https://doi.org/10.1016/s0169-5347(00)01994-7 (2000). Article  CAS  PubMed  PubMed Central  Google

Scholar  * Gao, F. _et al._ EasyCodeML: A visual tool for analysis of selection using CodeML. _Ecol. Evol._ 9, 3891–3898. https://doi.org/10.1002/ece3.5015 (2019). Article  PubMed  PubMed

Central  Google Scholar  * Ina, Y. Pattern of synonymous and nonsynonymous substitutions: An indicator of mechanisms of molecular evolution. _J. Genet._ 75, 91–115.

https://doi.org/10.1007/bf02931754 (1996). Article  CAS  Google Scholar  * Yang, Z. & Nielsen, R. Synonymous and nonsynonymous rate variation in nuclear genes of mammals. _J. Mol. Evol._

46, 409–418. https://doi.org/10.1007/pl00006320 (1998). Article  ADS  CAS  PubMed  Google Scholar  * Choudhuri, S. Fundamentals of molecular evolution. In _Bioinformatics for Beginners:

Genes, Genomes, Molecular Evolution, Databases and Analytical Tools_ (ed. Choudhuri, S.) 27–53 (Academic Press, 2014). https://doi.org/10.1016/b978-0-12-410471-6.00002-5. Chapter  Google

Scholar  * George, D. & Blieck, A. Rise of the earliest tetrapods: An early Devonian origin from marine environment. _PLoS One_ 6, e22136. https://doi.org/10.1371/journal.pone.0022136

(2011). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Zhang, F. & Deharveng, L. Systematic revision of Entomobryidae (Collembola) by integrating molecular and new

morphological evidence. _Zool. Scr._ 44, 298–311. https://doi.org/10.1111/zsc.12100 (2014). Article  Google Scholar  * Zhang, F., Sun, D., Yu, D. & Wang, B. Molecular phylogeny supports

S-chaetae as a key character better than jumping organs and body scales in classification of Entomobryoidea (Collembola). _Sci. Rep._ 5, 12471. https://doi.org/10.1038/srep12471 (2015).

Article  ADS  PubMed  PubMed Central  Google Scholar  * Chen, T., Sandmann, P., Schaefer, I. & Scheu, S. Neutral lipid fatty acid composition as trait and constraint in Collembola

evolution. _Ecol. Evol._ 7, 9624–9638. https://doi.org/10.1002/ece3.3472 (2017). Article  PubMed  PubMed Central  Google Scholar  * Ding, Y., Yu, D., Guo, W., Li, J. & Zhang, F.

Molecular phylogeny of Entomobrya (Collembola: Entomobryidae) from China: Colour pattern groups and multiple origins. _Insect Sci._ 26, 587–597. https://doi.org/10.1111/1744-7917.12559

(2018). Article  PubMed  Google Scholar  * Zhang, F. _et al._ Molecular phylogeny reveals independent origins of body scales in Entomobryidae (Hexapoda: Collembola). _Mol. Phylogenet. Evol._

70, 231–239. https://doi.org/10.1016/j.ympev.2013.09.024 (2014). Article  PubMed  Google Scholar  * Timmermans, M. J. T. N., Roelofs, D., Mariën, J. & van Straalen, N. M. Revealing

pancrustacean relationships: Phylogenetic analysis of ribosomal protein genes places Collembola (springtails) in a monophyletic Hexapoda and reinforces the discrepancy between mitochondrial

and nuclear DNA markers. _BMC Evol. Biol._ 8, 83. https://doi.org/10.1186/1471-2148-8-83 (2008). Article  CAS  PubMed  PubMed Central  Google Scholar  * Misof, B. _et al._ Phylogenomics

resolves the timing and pattern of insect evolution. _Science_ 346, 763–767. https://doi.org/10.1126/science.1257570 (2014). Article  ADS  CAS  PubMed  Google Scholar  * Beutel, R. G.,

Yavorskaya, M. I., Mashimo, Y. & Fukui, M. The phylogeny of Hexapoda (Arthropoda) and the evolution of megadiversity. _Proc. Arthropodan Embryol. Soc. Jpn._ 51, 1–15 (2017). Google

Scholar  * Carapelli, A., Liò, P., Nardi, F., Van der Wath, E. & Frati, F. Phylogenetic analysis of mitochondrial protein coding genes confirms the reciprocal paraphyly of Hexapoda and

Crustacea. _BMC Evol. Biol._ 7, 1–13. https://doi.org/10.1186/1471-2148-7-s2-s8 (2007). Article  Google Scholar  * D’Haese, C. A. Morphological appraisal of Collembola phylogeny with special

emphasis on Poduromorpha and a test of the aquatic. _Zool. Scr._ 32, 563–586. https://doi.org/10.1046/j.1463-6409.2003.00134.x (2003). Article  Google Scholar  * Magallón, S. A review of

the effect of relaxed clock method, long branches, genes, and calibrations in the estimation of angiosperm age. _Bot. Sci._ 92, 1–22. https://doi.org/10.17129/botsci.37 (2014). Article 

Google Scholar  * Pimentel, M., Escudero, M., Sahuquillo, E., Minaya, M. Á. & Catalán, P. Are diversification rates and chromosome evolution in the temperate grasses (Pooideae)

associated with major environmental changes in the Oligocene-Miocene?. _PeerJ_ 5, 1–29. https://doi.org/10.7717/peerj.3815 (2017). Article  CAS  Google Scholar  * Estep, M. C. _et al._

Allopolyploidy, diversification, and the Miocene grassland expansion. _PNAS_ 111, 15149–15154. https://doi.org/10.1073/pnas.1404177111 (2014). Article  ADS  CAS  PubMed  PubMed Central 

Google Scholar  * Strömberg, C. Decoupled taxonomic radiation and ecological expansion of open-habitat grasses in the Cenozoic of North America. _PNAS_ 102, 11980–11984.

https://doi.org/10.1073/pnas.0505700102 (2005). Article  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Fjellberg, A. _The Collembola of Fennoscandia and Denmark, Part II:

Entomobryomorpha and Symphypleona_ (Koninklijke Brill NV, 2007). Book  Google Scholar  * Moreau, C. S., Bell, C., Vila, R., Archibald, S. & Pierce, N. Phylogeny of the ants:

Diversification in the age of angiosperms. _Science_ 312, 101–104. https://doi.org/10.1126/science.1124891 (2006). Article  ADS  CAS  PubMed  Google Scholar  * Peters, R. S. _et al._

Evolutionary history of the Hymenoptera. _Curr. Biol._ 27, 1013–1018. https://doi.org/10.1016/j.cub.2017.01.027 (2017). Article  CAS  PubMed  Google Scholar  * Prinzing, A. & Woas, S.

Habitat use and stratification of Collembola and oribatid mites. In _Arthropods of Tropical Forests: Spatio-temporal Dynamics and Resource Use in the Canopy_ (eds Basset, Y. _et al._)

271–281 (Cambridge University Press, 2003). Google Scholar  * Satta, Y., Ishiwa, H. & Chigusa, S. I. Analysis of nucleotide substitutions of mitochondrial DNAs in _Drosophila

melanogaster_ and its sibling species. _Mol. Biol. Evol._ 4, 638–650. https://doi.org/10.1093/oxfordjournals.molbev.a040464 (1987). Article  CAS  PubMed  Google Scholar  * Moriyama, E. N.

& Powell, J. R. Synonymous substitution rates in _Drosophila_: Mitochondrial versus nuclear genes. _J. Mol. Evol._ 45, 378–391. https://doi.org/10.1007/pl00006243 (1997). Article  ADS 

CAS  PubMed  Google Scholar  * Lawrie, D. S., Messer, P. W., Hershberg, R. & Petrov, D. A. Strong purifying selection at synonymous sites in _D. melanogaster_. _PLoS Genet._ 9, 33–40.

https://doi.org/10.1371/journal.pgen.1003527 (2013). Article  CAS  Google Scholar  * Nachman, M. W. Deleterious mutations in animal mitochondrial DNA. _Genetica_ 102(103), 61–69.

https://doi.org/10.1007/978-94-011-5210-5_6 (1998). Article  PubMed  Google Scholar  * Meiklejohn, C. D., Montooth, K. L. & Rand, D. M. Positive and negative selection on the

mitochondrial genome. _Trends Genet._ 23, 259–263. https://doi.org/10.1016/j.tig.2007.03.008 (2007). Article  CAS  PubMed  Google Scholar  * Stewart, J. B. _et al._ Strong purifying

selection in transmission of mammalian mitochondrial DNA. _PLoS Biol._ 6, 63–71. https://doi.org/10.1371/journal.pbio.0060010 (2008). Article  CAS  Google Scholar  * Ponnuswamy, P. K.,

Prabhakaran, M. & Manavalan, P. Hydrophobic packing and spatial arrangement of amino acid residues in globular proteins. _Biochim. Biophys. Acta_ 623, 301–316.

https://doi.org/10.1016/0005-2795(80)90258-5 (1980). Article  CAS  PubMed  Google Scholar  * Gromiha, M. M., Oobatake, M., Kono, H., Uedaira, H. & Sarai, A. Relationship between amino

acid properties and protein stability: Buried mutations. _J. Protein Chem._ 18, 565–578. https://doi.org/10.1023/a:1020603401001 (1999). Article  CAS  PubMed  Google Scholar  * Kumar, S.,

Stecher, G., Li, M., Knyaz, C. & Tamura, K. MEGA X: Molecular evolutionary genetics analysis across computing platforms. _Mol. Biol. Evol._ 35, 1547–1549.

https://doi.org/10.1093/molbev/msy096 (2018). Article  CAS  PubMed  PubMed Central  Google Scholar  * Xia, X. DAMBE7: New and improved tools for data analysis in molecular biology and

evolution. _Mol. Biol. Evol._ 35, 1550–1552. https://doi.org/10.1093/molbev/msy073 (2018). Article  CAS  PubMed  PubMed Central  Google Scholar  * Castoe, T. A., Jiang, Z. J., Gu, W., Wang,

Z. O. & Pollock, D. D. Adaptive evolution and functional redesign of core metabolic proteins in snakes. _PLoS One_ 3, e2201. https://doi.org/10.1371/journal.pone.0002201 (2008). Article

  ADS  CAS  PubMed  PubMed Central  Google Scholar  * Xia, X. & Lemey, P. Assesing substitution saturation with DAMBE. In _Phylogenetic Handbook: A Practical Approach to Phylogenetic

Analysis and Hypothesis Testing_ (eds Lemey, P. _et al._) 611–626 (Cambridge University Press, 2009). Google Scholar  * Xia, X., Hafner, M. S. & Sudman, P. D. On transition bias in

mitochondrial genes of pocket gophers. _J. Mol. Evol._ 43, 32–40. https://doi.org/10.1007/bf02352297 (1996). Article  ADS  CAS  PubMed  Google Scholar  * Bouckaert, R. _et al._ BEAST 2: A

software platform for Bayesian evolutionary analysis. _PLoS Comput. Biol._ 10, e1003537. https://doi.org/10.1371/journal.pcbi.1003537 (2014). Article  CAS  PubMed  PubMed Central  Google

Scholar  * Deharveng, L. & Bedos, A. Diversity of terrestrial invertebrates in subterranean habitats. In _Cave Ecology, Ecological Studies_ (eds Moldovan, O. T. _et al._) 107–172

(Springer International Publishing, 2018). Chapter  Google Scholar  * Letunic, I. & Bork, P. Interactive Tree Of Life (iTOL): An online tool for phylogenetic tree display and annotation.

_Bioinformatics_ 23, 127–128. https://doi.org/10.1093/bioinformatics/btl529 (2007). Article  CAS  PubMed  Google Scholar  * Xiong, Y., Gao, Y., Yin, W. & Luan, Y. Molecular phylogeny of

Collembola inferred from ribosomal RNA genes. _Mol. Phylogenet. Evol._ 49, 728–735. https://doi.org/10.1016/j.ympev.2008.09.007 (2008). Article  CAS  PubMed  Google Scholar  * Sun, X. _et

al._ Phylomitogenomic analyses on collembolan higher taxa with enhanced taxon sampling and discussion on method selection. _PLoS One_ 15, e0230827.

https://doi.org/10.1371/journal.pone.0230827 (2020). Article  CAS  PubMed  PubMed Central  Google Scholar  * Cucini, C. _et al._ Re-evaluating the internal phylogenetic relationships of

Collembola by means of mitogenome data. _Genes_ 12, 44. https://doi.org/10.3390/genes12010044 (2021). Article  CAS  Google Scholar  * Leo, C., Carapelli, A., Cicconardi, F., Frati, F. &

Nardi, F. Mitochondrial genome diversity in Collembola: Phylogeny, dating and gene order. _Diversity_ 11, 169. https://doi.org/10.3390/d11090169 (2019). Article  CAS  Google Scholar  *

Miller, M. A. _et al._ A RESTful API for access to phylogenetic tools via the CIPRES Science Gateway. _Evol. Bioinform._ 11, 43–48. https://doi.org/10.4137/ebo.s21501 (2015). Article  CAS 

Google Scholar  * Rambaut, A., Drummond, A., Xie, D., Baele, G. & Suchard, M. A. Posterior summarization in bayesian phylogenetics using Tracer 1.7. _Syst. Biol._ 67, 901–904.

https://doi.org/10.1093/sysbio/syy032 (2018). Article  CAS  PubMed  PubMed Central  Google Scholar  * Maddison, W. P. & Maddison, D. R. Mesquite: A modular system for evolutionary

analysis. http://www.mesquiteproject.org (2019). * Lewis, P. O. A likelihood approach to estimating phylogeny from discrete morphological character data. _Syst. Biol._ 50, 913–925.

https://doi.org/10.1080/106351501753462876 (2001). Article  CAS  PubMed  Google Scholar  * Pagel, M., Meade, A. & Barker, D. Bayesian estimation of ancestral character states on

phylogenies. _Syst. Biol._ 53, 673–684. https://doi.org/10.1080/10635150490522232 (2004). Article  PubMed  Google Scholar  * Jordan, F. M., Gray, R. D., Greenhill, S. J. & Mace, R.

Matrilocal residence is ancestral in Austronesian societies. _Proc. R. Soc. B_ 276, 1957–1964. https://doi.org/10.1098/rspb.2009.0088 (2009). Article  PubMed  PubMed Central  Google Scholar

  * Yang, Z. PAML 4: Phylogenetic analysis by maximum likelihood. _Mol. Biol. Evol._ 24, 1586–1591. https://doi.org/10.1093/molbev/msm088 (2007). Article  CAS  PubMed  Google Scholar  *

Yang, Z. & Nielsen, R. Codon-substitution models for detecting molecular adaptation at individual sites along specific lineages. _Mol. Biol. Evol._ 19, 908–917.

https://doi.org/10.1093/oxfordjournals.molbev.a004148 (2002). Article  CAS  PubMed  Google Scholar  * Wickham, H. _ggplot2: Elegant Graphics for Data Analysis_ (Springer, 2016). Book  Google

Scholar  * Woolley, S., Johnson, J., Smith, M. J., Crandall, K. A. & Mcclellan, D. A. TreeSAAP: Selection on amino acid properties using phylogenetic trees. _Bioinform. Appl. Note_ 19,

671–672. https://doi.org/10.1093/bioinformatics/btg043 (2003). Article  CAS  Google Scholar  * McClellan, D. A. & Ellison, D. D. Assessing and improving the accuracy of detecting protein

adaptation with the TreeSAAP analytical software. _Int. J. Bioinform. Res. Appl._ 6, 120–133. https://doi.org/10.1504/ijbra.2010.032116 (2010). Article  CAS  PubMed  Google Scholar 

Download references ACKNOWLEDGEMENTS The authors thank the National Research Foundation (NRF) Competitive Programme for Rated Researcher grant to B.v.V (NRF CPRR: grant number 138016), the

University of Johannesburg funding to B.v.V. and GES 4.0 fellowship awarded to D.M.M. and D.C.M., and the South African (NRF)/Russia (RFBR) Joint Science and Technology Research

Collaboration project no. 19-516-60002 and no. 118904 (NRF) awarded to M.P. and C.JS., respectively. The authors wish to thank the analyses platform and bioinformatics support provided by

the Centre for High Performance Computing (CHPC) in Cape Town, South Africa. We appreciate the constructive criticism received by the two anonymous reviewers which improved the manuscript.

AUTHOR INFORMATION AUTHORS AND AFFILIATIONS * Centre for Ecological Genomics and Wildlife Conservation, Department of Zoology, University of Johannesburg, Auckland Park, 2006, South Africa

Daniela M. Monsanto, Devon C. Main, Arsalan Emami-Khoyi, Shilpa P. Parbhu, Peter R. Teske & Bettine Jansen van Vuuren * Department of Biological Sciences, University of Cape Town,

Rondebosch, 7701, South Africa Charlene Janion-Scheepers * Iziko South African Museum, Gardens, Cape Town, 8001, South Africa Charlene Janion-Scheepers * Institut de Systématique, Evolution,

Biodiversité (ISYEB)-UMR 7205-CNRS, UPMC, EPHE, Muséum National d’Histoire Naturelle (MNHN), Sorbonne Université, 75231, Paris, France Louis Deharveng & Anne Bedos * Department of

Zoology and Ecology, Institute of Biology and Chemistry, Moscow State Pedagogical University, Moscow, 119991, Russia Mikhail Potapov * Senckenberg Museum of Natural History, 60325, Görlitz,

Germany Mikhail Potapov * School of Natural Sciences, Macquarie University, Sydney, NSW, 2109, Australia Johannes J. Le Roux Authors * Daniela M. Monsanto View author publications You can

also search for this author inPubMed Google Scholar * Devon C. Main View author publications You can also search for this author inPubMed Google Scholar * Charlene Janion-Scheepers View

author publications You can also search for this author inPubMed Google Scholar * Arsalan Emami-Khoyi View author publications You can also search for this author inPubMed Google Scholar *

Louis Deharveng View author publications You can also search for this author inPubMed Google Scholar * Anne Bedos View author publications You can also search for this author inPubMed Google

Scholar * Mikhail Potapov View author publications You can also search for this author inPubMed Google Scholar * Shilpa P. Parbhu View author publications You can also search for this

author inPubMed Google Scholar * Johannes J. Le Roux View author publications You can also search for this author inPubMed Google Scholar * Peter R. Teske View author publications You can

also search for this author inPubMed Google Scholar * Bettine Jansen van Vuuren View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS D.M.M.,

A.E.-K., and B.v.V. designed the research; D.M.M. and D.C.M. analysed data; D.M.M., C.J.-S., L.D., A.B., M.P., and S.P.P. compiled and corrected the ecological life form Supplementary Table;

all authors contributed to the writing of the manuscript. CORRESPONDING AUTHOR Correspondence to Bettine Jansen van Vuuren. ETHICS DECLARATIONS COMPETING INTERESTS The authors declare no

competing interests. ADDITIONAL INFORMATION PUBLISHER'S NOTE Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

SUPPLEMENTARY INFORMATION SUPPLEMENTARY INFORMATION 1. SUPPLEMENTARY INFORMATION 2. RIGHTS AND PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons Attribution 4.0

International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the

source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's

Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not

permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit

http://creativecommons.org/licenses/by/4.0/. Reprints and permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Monsanto, D.M., Main, D.C., Janion-Scheepers, C. _et al._ Mitogenome selection in

the evolution of key ecological strategies in the ancient hexapod class Collembola. _Sci Rep_ 12, 14810 (2022). https://doi.org/10.1038/s41598-022-18407-1 Download citation * Received: 12

December 2021 * Accepted: 10 August 2022 * Published: 31 August 2022 * DOI: https://doi.org/10.1038/s41598-022-18407-1 SHARE THIS ARTICLE Anyone you share the following link with will be

able to read this content: Get shareable link Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing

initiative