Cryptic diversity and phylogeographic patterns of Deto echinata (Isopoda: Detonidae) in southern Africa

Recent phylogeographic studies of poorly-dispersing coastal invertebrates in highly biodiverse regions have led to the discovery of high levels of cryptic diversity and complex phylogeographic patterns that suggest isolation, geological, and ecological processes have shaped their biodiversity. Studies of southern African coastal invertebrates have uncovered cryptic diversity for various taxa and phylogeographic patterns that, although sharing some similarities across taxa, do differ. These findings underscore the need for additional studies to better understand the biodiversity levels, distributional patterns, and processes responsible for producing coastal biodiversity in that region. The coastal isopod Deto echinata is of particular interest, as its complex taxonomic history, poor dispersal capabilities, and broad geographic distribution suggest the potential for cryptic diversity. We use mitochondrial and nuclear sequences to characterize D. echinata individuals from localities ranging from northern Namibia to Glentana, about 2,500 km along the coastline on the south coast of South Africa. These are used to assess whether D. echinata harbors cryptic genetic diversity and whether phylogeographic distributional patterns correlate with those previously documented for other coastal isopods in the region. Analysis of mitochondrial and nuclear sequences revealed two deeply-divergent lineages that exhibit a distributional break in the Cape Peninsula region. These findings suggest D. echinata is a cryptic species complex in need of taxonomic revision and highlight the need for further taxonomic and phylogeographic studies of similarly poorly-dispersing coastal invertebrates in southern Africa.

Previous molecular studies of coastal isopods with complex taxonomic histories have shown them to harbor cryptic diversity and complex phylogenetic patterns (Hurtado, Mateos & Santamaria, 2010;Hurtado, Lee & Mateos, 2013;Santamaria et al., 2013;Santamaria, Mateos & Hurtado, 2014;Santamaria & Koch, 2023).Drift-line isopods in the genus Deto (Guérin-Méneville, 1836) are one potential taxonomic group of interest.This genus is currently considered to have four well-established species, D. aucklandiae Thomson 1879 and D. bucculenta Nicolet 1849 from New Zealand and its outlying islands; D. marina Chilton 1885 from southern Australia; and D. echinata (Guérin-Méneville, 1836) from southern Africa (with an outlying population reported from St Paul Island in the Indian Ocean) (Schmalfuss, 2003).A fifth species, D. whitei (Kinahan, 1859), is of doubtful status, as its type location remains unknown and it has not been reported since its original and extremely cursory description by Kinahan (1859).This has led several authors (Budde-Lund, 1885;Chilton, 1915;Schmalfuss, 2003) to suggest that it is a synonym of D. echinata; indeed, the original author himself remarked on its close similarly to that species.
Deto echinata was first described by Guérin-Méneville (1836) and is readily distinguished by the pairs of long dorsal 'horns' projecting from each thoracic segment.The type locality of the species remains obscure, since the original description simply gives its type locality as 'd'Orient' or 'in the east' (Guérin-Méneville, 1836).Moreover, the original specimens cannot be traced.The subsequent taxonomic history of D. echinata is long and confused.Krauss (1843) provided the first unambiguous South African record, which he collected from the sea-shore in Table Bay.Budde-Lund (1885) also recorded D. echinata from the region, as well as another species, D. acinosa, which he described as new.Chilton (1915) provided a full review of the genus as it stood at that time and retained both South African species, which he distinguished primarily by the sizes of the dorsal projections, which he described as long in D. echinata, but as reduced to tubercles in D. acinosa.The same distinction was maintained by Panning (1924) with respect to specimens from Namibia, but Barnard (1932) concluded that D. acinosa simply represented 'not fully grown, but not necessarily sexually immature' specimens and amalgamated all southern African specimens under the name D. echinata (although he still depicted the acinosa 'form').Subsequent authors have referred to the species simply as D. echinata and have not recognized sub-specific taxa (Kensley, 1978;Coleman & Leistikow, 2001;Branch et al., 2016).This decision appears to be supported by Glazier, Clusella-Trullas & Terblanche (2016) who showed that horn length in this species not only varies with sex and body size, but is also strongly influenced by nutritional condition.
An outlying population of D. echinata is also known from Saint Paul Island, in the Southern Indian Ocean (33 • S, 77 • E), midway between South Africa and Australia.That population was originally described as a separate species, D. armata, by Budde-Lund (1906).Chilton (1915) continued to recognize this as a separate species, although noting its similarity to D. echinata.However, both Barnard (1932) and Coleman & Leistikow (2001) regarded it as a junior synonym of D. echinata and that remains its current status.
Deto echinata is an air-breathing species found in vast numbers in the upper intertidal and supratidal regions of rocky shores and has a reported range from northern Namibia to about Knysna on the south coast of South Africa (Fig. 1, Barnard, 1932;Kensley, 1978).It occurs along with similar-looking (but spineless and only distantly-related), members of the genus Ligia (family Ligiidae), all species feeding on drift algae deposited along the drift-line of rocky shores.Despite the extraordinary abundance of D. echinata at some sites, its complex taxonomic history, and the range overlap with Ligia spp., for which cryptic diversity has previously been reported (see Greenan, Griffiths & Santamaria, 2018), there have been no previous publications on phylogeography or cryptic diversity of this isopod.
In this study, we use mitochondrial and nuclear markers to characterize D. echinata individuals from diverse localities across its range in Namibia and South Africa to determine whether cryptic genetic diversity exists within these.We also examine the distributional patterns of the genetic diversity of D. echinata and compare them to previous phylogeographic patterns for coastal isopods in the region.

Specimen collections
Deto individuals were hand-collected from the upper intertidal zone at 16 coastal localities chosen as to span the known geographic range of Deto echinata as well as reported phylogeographic breaks in the region (Fig. 1).Localities span from Rocky Point in northern Namibia to Glentana, on the south coast of South Africa, a coastal distance of ∼2,500 km.Sampling attempts in some other localities in the southern coastline of South Africa east of L'Agulhas (e.g., Arniston, De Hoop, and Mossel Bay) were not successful, as no specimens could be found, despite suitable habitat being searched.Detailed information on sampling localities included in this study is provided in Table 1.All samples were field-preserved and stored in 95% ethanol pending molecular analyses.Field collections were carried out under Scientific Collection Permit RES2017/53 issued to CG jointly by the South African Departments of Environmental Affairs and of Agriculture, Forestry and Fisheries.

Sequence alignments, phylogenetic analyses, and estimation of molecular divergence
Sequences produced in this study were sorted by gene, with each dataset then aligned independently using the MAFFT algorithm (Katoh & Standley, 2013;Katoh, Rozewicki & Yamada, 2019) as implemented in the GUIDANCE2 Server (Sela et al., 2015) with all settings as default except the number of bootstrap replicates, which was set to 100.
Positions with an alignment score <1.00 in the final alignment were considered poorly aligned and excluded from posterior analyses.Mitochondrial genes were concatenated into a single alignment in SequenceMatrix v1.6.7 (Vaidya, Lohman & Meier, 2011).The nuclear 28S rDNA gene was not incorporated into this dataset, given the low levels of variation observed (see Results).Instead, we estimated relationships between 28S haplotypes using the cladogram estimation algorithm of (Templeton, Crandall & Sing, 1992), as implemented by PopART v1.7 (Leigh & Bryant, 2015).
We used jModeltest v2.1.1 (Darriba et al , 2012) to determine the most appropriate model of DNA substitution for each mitochondrial gene fragment and the mitochondrial concatenated dataset by evaluating likelihood scores of 1,624 candidate models on a fixed BioNJ-JC tree, under the Akaike Information Criterion (AIC), corrected AIC (AICc), and the Bayesian Information Criterion (BIC).Selected models were used in phylogenetic reconstructions, unless the chosen model was not available in the software being used, or if the chosen model called for the joint estimation of and I parameters.In case of the former, we used the next more complex model available in the software, while in the case of the latter, we used the simpler .
We attempted to identify an appropriate outgroup to root our phylogenetic reconstructions by incorporating publicly available sequences for other Deto species; however, the only available data at the time of analyses in May of 2023 were COI sequences for D. marina (GenBank accession numbers: KR424585, KR424586, and EU364625).Thus, we aligned these sequences with the D. echinata COI sequences produced herein and conducted preliminary phylogenetic reconstructions in RAxML v8.2.12 (Stamatakis, 2014), as described below.The reconstructions recovered a split between two moderately supported D. echinata clades when rooted with D. marina (BS >75) with the topology of the tree remaining unchanged upon using a mid-point root.Thus, all posterior phylogenetic analyses were rooted using each identified clade to root the other clade as implemented by both Mateos et al. (2012) and Santamaria & Koch (2023).
Phylogenetic reconstructions were carried out on the concatenated mitochondrial dataset using both Maximum Likelihood (ML) and Bayesian Inference (BI) under two partitioning schemes (i.e., unpartitioned, partitioned by gene).ML searches were carried out in RAxML v8.2.12 (Stamatakis, 2014) and consisted of 1,000 bootstrap replicates followed by a thorough ML search under the GTR + model run under the Rapid Bootstrap Algorithm (Stamatakis, Hoover & Rougemont, 2008) with all other settings as default.For each search, we estimated a majority-rule consensus of all bootstrap replicates using the SumTrees command of DendroPy v4.1.0(Sukumaran & Holder, 2010).BI searches were carried out in MrBayes v3.2.5 (Ronquist & Huelsenbeck, 2003) and Phycas v2.2.0 (Lewis, Holder & Swofford , 2015).Searches in MrBayes consisted of two simultaneous searches of four chains run for 20× 10 6 generations sampled every 5,000th generation, while Phycas searches consisted of a single search of 2× 10 6 generations sampled every 50th generation.We determined whether Bayesian analyses had reached convergence if the average standard deviation of the split frequencies of independent runs was stable and close to zero, and if the Effective Sample Size (ESS) for posterior probabilities exceeded 200 when evaluated in Tracer v.1.7 (Rambaut et al., 2018).We estimated node support values by discarding all samples prior to stationarity (10-25% of sampled trees) and calculating a majority-rule consensus tree using the SumTrees command of DendroPy v4.1.0(Sukumaran & Holder, 2010).
ASAP analyses were carried out separately on the COI dataset and on the concatenated mitochondrial dataset using the ASAP online server (https://bioinfo.mnhn.fr/abi/public/asap/) using the Kimura 2-Parameter (K2P) nucleotide evolution model, a ts/tv ratio of 2, and all other settings as default.Individuals in the concatenated mitochondrial dataset with missing data were removed prior to analyses.
PTP analyses were conducted on the consensus trees produced by phylogenetic reconstructions carried out in RAxML and MrBayes.PTP analyses were carried out under the Maximum Likelihood implementation of PTP (Zhang et al., 2013) using 500,000 MCMC iterations, a random seed, a burn-in of 0.10, and a thinning value of 100.GMYC analyses were conducted on ultrametric trees inferred in BEAST v2.1.3(Bouckaert et al., 2014) using three different approaches: (a) assuming a constant rate of evolution and speciation under a Yule process (Yule, 1925;hereafter Constant + Yule, Gernhard, 2008); (b) assuming a relaxed clock and speciation under a Yule process (hereafter Relaxed + Yule), and (c) under a coalescent model of speciation assuming a constant population size (Kingman, 1982;hereafter Coalescent).All BEAST searches were carried out for 10× 10 6 generations sampled every 1,000th generation using the GTR + model of nucleotide substitution.Searches were evaluated for convergence using the criteria previously described for Bayesian phylogenetic reconstructions.Trees were summarized using TreeAnnotator v1.8.2 (https://beast.community/treeannotator)with 10% of trees discarded as burn-in and edges set using the mean-age option.Resulting ultrametric trees were analyzed in R using GMYC approach implemented by the 'splits' package (http://r-forge.r-project.org/projects/splits/).

Morphological comparisons
A subset of specimens were observed and drawn using a Wild binocular microscope fitted with a 1.25X camera lucida with photographs taken using a Nikon D3100 DSLR camera fitted with a 95 mm macro lens.Individuals were sexed based on the presence of conspicuous dorsal horns in males and of a marsupium in adult females.Body lengths were measured from the anterior of the cephalon to the posterior tip of the pleotelson.

RESULTS
Our final mitochondrial concatenated dataset consisted of 76 Deto individuals from 16 localities across the coastlines of both Namibia and South Africa (COI = 71 individuals, 16S rDNA = 55 individuals, 12S rDNA = 60 individuals).The final alignment of mitochondrial genes was 1,593-bp long after the exclusion of 26 nucleotide positions: 15-bp from the 16S rDNA and 11-bp from the 12S rDNA alignments.The final 28S rDNA alignment was 589-bp long after the exclusion of four nucleotide positions and included 30 individuals.All sequences produced in this study have been deposited in GenBank under accession numbers OQ852632-OQ852680, OQ870135-OQ870160, and OQ870100-OQ870134 (Table 1) and in the Barcode of Life Database (BOLD; BOLD BINs ADE2360 and ADV7260; specimens DECAC001-23 to DECAC076-23).Alignments with poorly aligned positions are provided as Dataset S1 and S2.

Phylogenetic reconstructions
Preliminary phylogenetic reconstructions based on COI sequences including D. marina identified a moderately-supported basal split between two large D. echinata clades: a west coast ''Namibia-Cape Town'' clade (yellows and reds in all figures; Bootstrap support: 99 in Fig. 2), and a south coast ''False Bay-Mossel Bay'' clade (blues in all figures; Bootstrap support: 100).The ''Namibia-Cape Town'' clade included all individuals from localities in Namibia (A1-3), all those in South African samples collected between Jacobsbaai to Bakoven (A4-8), and one specimen collected from Simon's Town in False Bay (B2).The ''False Bay-Mossel Bay'' clade included all specimens collected in Kommetjie, South Africa (B1) the remaining specimens collected from Simon's Town in False Bay (B2) and all specimens collected to the east of this along the South African south coast (B3-B8).
Give the high level of divergence between D. marina and the in-group (Table 2), and the lack of sequence data for D. marina beyond COI sequences, we conducted our phylogenetic analyses on the concatenated mitochondrial dataset without D. marina, instead assuming a mid-point rooting scheme.All analyses identified the clades described above, however, the reciprocal monophyly of each clade was highly supported in all analyses (BS = 100 in all ML searches, MPP = 100 in all BI searches, Fig. 2).These analyses produced slightly improved resolution within the ''Namibia-Cape Town'' clade where we recovered the monophyly of all specimens collected in Namibian localities (A1-3) and one specimen collected in Simon's Town in South Africa (B2) (BS = 76-85; MPP = 81-99).Relationships within the ''False Bay-Mossel Bay'' clade remained poorly resolved.

Molecular species delimitation analyses
Regardless of the dataset analyzed, ASAP analyses identified two putative species as the most appropriate hypothesis for the number of species present among our D. echinata specimens (Fig. 2, bars A and B).For the COI dataset, the two species hypothesis produced an ASAP-score of 1.5, with a p-value of 0.381 (rank = 2), a W value of 1.25 ×10 −4 (rank = 1), and a threshold distance of 0.041458.For the concatenated mitochondrial dataset, ASAP identified the two species hypothesis as the highest ranked solution with an ASAP score of 2, a p-value of 0.244 (rank = 1), a W value of 1.88 ×10 −4 (rank = 3), and a threshold distance of 0.028554.PTP and GMYC analyses identified two to four putative species clusters (Fig. 2).PTP analyses identified two putative species clusters when analyzing the phylogeny produced in RAxML assuming a single partition (Fig. 2, bar C) and four clusters for the phylogeny produced when assuming partitioning by genes (Fig. 2, bar D).Meanwhile, PTP analyses identified three putative species clusters for the phylogenies produced in MrBayes, regardless of partitioning used.For GMYC analyses, the Relaxed + Yule analysis identified two putative species, the Coalescent analysis identified three putative species, and the Strict Clock + Yule analysis identified four.
Results of MSDAs were largely congruent both in the number of putative species clusters identified and the assignment of individuals to clusters with differences between analyses due to splitting of larger clusters found in more inclusive analyses.Both ASAP analyses, the PTP analysis based on the unpartitioned ML phylogenetic reconstruction, and the GMYC Relaxed Clock + Yule analysis identified two putative species clusters.Each of these clusters corresponded to one of the major clades recovered in our phylogenetic reconstructions, with all ''Namibia-Cape Town'' clade individuals grouped in a single putative species and all ''False Bay-Mossel Bay'' clade individuals placed in a separate putative species (Fig. 2).Both PTP analyses of MrBayes phylogenetic reconstructions recovered a species cluster consisting of all ''Namibia-Cape Town'' individuals and two species consisting of ''False Bay-Mossel Bay'' individuals, one consisting of a single individual from Kalk Bay (B03) and the other containing the rest of the members of the clade.The remaining analyses identified a single species that consisted of all ''False Bay-Mossel Bay'' individuals, differing from the two species cluster solution by the splitting of ''Namibia-Cape Town'' individuals into two to three species depending on the analysis (Fig. 2, bars D, H, and I).

Morphological comparisons
Although we made initial observations on the morphological appearance of specimens from the two recognized clades, we were unable to detect any consistent features which could be used to separate the two 'species'.
The two D. echinata lineages reported herein exhibit amongst lineage COI K2P divergences that exceed 6.0% (Table 2).These divergences greatly exceed the broad cut-off suggested to delineate between within and amongst species genetic distances for metazoa (Hebert et al., 2003) and crustacean species (Matzenda Silva et al., 2011;Raupach et al., 2015).Indeed, the genetic distances amongst the lineages reported match, or exceed, those reported for other valid species of coastal isopods (e.g., Hurtado, Lee & Mateos, 2013;Santamaria et al., 2013;Santamaria et al., 2017;Hurtado et al., 2018), including recently described cryptic species (e.g., Santamaria, 2019).The two lineages also harbored different and unique 28S rDNA haplotypes, suggesting that they also exhibit differences in nuclear loci.Thus, our findings suggest D. echinata is a cryptic species complex in need of taxonomic revision.
Our exploratory MSDAs suggest that D. echinata represents at least two species, as most MSDAs identified the members of each highly divergent genetic lineage recovered in phylogenetic reconstructions as corresponding to separate species and BOLD recognized two separate BINS in our dataset.Thus, we suggest Deto populations from Namibia to Bakoven, South Africa are likely to represent one species, with populations from Kommetjie to Glentana, South Africa representing the other.
Despite convincing genetic evidence of D. echinata being a cryptic species complex, we did not observe diagnostic differences between these putative species in three morphological traits used in Deto taxonomy such as horn length, horn shape, and positioning of pleopods.However, horn length and shape are known to differ not only between male and females, but also between individuals of different sizes, and between individuals differing in body condition (Glazier, Clusella-Trullas & Terblanche, 2016).Thus, the absence of differences between the two clades in our preliminary morphological evaluations may be due to the confounding effects of sex, size, and/or body condition among the collected samples.These confounding effects may also account for previous taxonomic confusion in D. echinata, such as the erection and subsequent synonymization of D. acinosa based on differences in horn size.Given the complexity of this situation, we thus have elected not to present our inconclusive preliminary morphological observations here.Rather, we suggest that a separate comprehensive taxonomic work is needed to determine whether diagnostic morphological differences exist between the D. echinata lineages herein reported.
We suggest that future taxonomic revisions consider the two clades recovered in our phylogenetic analyses as hypothetical or putative species.Future taxonomic work should also incorporate outlying D. echinata populations, as these isopods are reputed to occur in Saint Paul Island in the Southern Indian Ocean (33 • S, 77 • E), midway between South Africa and Australia.Given its distance to the localities included in this study, it is likely that this outlying population represents another unique genetic lineage, different from those reported herein.

Phylogeographic patterns
Molecular studies on coastal invertebrates in South Africa have uncovered complex phylogeographic patterns not congruent across taxa, with related species exhibiting differing patterns (e.g., Teske et al., 2006;Reynolds, Matthee & Heyden, 2014;Mmonwa et al., 2015;Baldanzi et al., 2016;Greenan, Griffiths & Santamaria, 2018;von der Heyden, Mbongwa & Hui, 2020).Despite contrasting patterns, most of these poorly dispersing coastal invertebrates exhibit strong phylogeographic breaks around the Cape of Good Hope, which is regarded as the most significant biogeographic transition zone around the entire South African coastline (Griffiths et al., 2010).Greenan, Griffiths & Santamaria (2018) saw a break in the distribution of Ligia lineages, with a major lineage distributed from Ganzekraal, South Africa (A05 herein) to Luderitz, Namibia (A03) and another lineage being distributed from Kommetjie (B01) to L'Agulhas in South Africa.von der Heyden, Mbongwa & Hui (2020) also found a break in the distribution of lineages for two Excirolana species between localities in the west of Cape Point and those in False Bay and eastward, with the phylogeographic break occurring in the region between Yzerfontein and Simon's Town (B02) in False Bay.Meanwhile, Teske et al. (2006) identified two major and non-overlapping lineages within the sphaeromatid isopod Exosphaeroma hylecoetes from the regions around the Cape Peninsula: a south-western lineage that included individuals collected in estuarine localities between Cape Agulhas and the Cape of Good Hope, and a western lineage from localities west of the Cape of Good Hope.The two highly divergent genetic D. echinata lineages uncovered in this study are no exception, as the two lineages we uncover exhibit a strong phylogeographic break in the region, with the ''Namibia-Cape Town'' and the ''False Bay-Mossel Bay'' exhibiting a break in distribution somewhere in the area Bakoven (A07) and Kommetjie (B01).
Despite the strong phylogeographic break we report herein and the poor dispersal capabilities of D. echinata, we did observe evidence suggesting that rare dispersal events may take place across the Cape of Good Hope.A single D. echinata specimen collected in Simon's Town (B02), a location in False Bay east of the phylogeographic break, harbored mitochondrial haplotypes that placed it in the ''Namibia-Cape Town'' clade.Interestingly, this individual shared 28S rDNA haplotypes with individuals from the ''False Bay-Mossel Bay'' clade.These patterns suggest this individual from Simon's Town may represent a hybrid, or a recent migrant harboring an ancestral polymorphism at the 28S rDNA gene.
Phylogeographic patterns of D. echinata exhibit some important differences to those exhibited by other coastal isopods in the region.For instance, D. echinata populations do not appear to exhibit the high levels of population differentiation and structuring reported for Ligia isopods (Greenan, Griffiths & Santamaria, 2018) and Tylos granulatus (Mbongwa et al., 2019;Bezuidenhout et al., 2021).Whereas genetic divergences amongst populations within lineages in Ligia from southern Africa ranged exceed 5.0% COI K2P in many instances (Greenan, Griffiths & Santamaria, 2018), COI K2P divergences amongst localities of D. echinata within the same lineage rarely exceed 2.5%.The exception to this pattern being the relatively higher levels of differentiation seen amongst Namibia populations and South Africa populations within the ''Namibia-Cape Town'' lineage (1.9-3.1% COI K2P) indicating at least a potential effect of distance in the isolation and genetic differentiation of Deto populations in southern Africa.
The presence of phylogeographic breaks in the Cape Peninsula region for these taxa, appear to suggest that the area's geological history may be the primary driver of the distribution of genetic diversity for coastal organisms in the area (Toms et al., 2014).
Subtle differences in the location of phylogeographic breaks in this region; however, may reflect differences in life history, evolutionary history, and ecological needs amongst taxa (Pelc, Warner & Gaines, 2009).The Cape Peninsula region is known to exhibit ecological gradients in both biotic and abiotic factors, as a result of the varying influences of the colder Atlantic waters and warmer Indian Ocean to the east and west of the Cape Peninsula respectively (Bustamante et al., 1995;Demarcq, Barlow & Shillington, 2003).Indeed, Baldanzi et al. (2016) suggested that the unusual phylogeographic break seen in the amphipod Talorchestia (now Africorchestia) capensis in the Cape region may be due to species-specific environmental conditions and requirements.Differences between the biology of organisms may thus help explain the location of phylogeographic breaks reported for different organisms in the Cape region.For instance, the distribution of lineages of air-breathing organisms that rarely submerge, such as Deto, Ligia, Tylos, and Talorchestia may be more likely influenced by terrestrial climatic conditions, while the distribution of underwater breathing organisms, such as Exosphaeroma, are more likely influenced by marine environmental conditions.Relatedly, differences in climatic conditions as well as climate oscillations may explain the differences in divergence and divergence time estimates reported for coastal organisms in the Cape region (Teske et al., 2007;Teske et al., 2009).In the case of Deto, the moderate level of divergence between the two major clades uncovered suggest their origin likely predate the Last Glacial Maxima.Rough estimates of the divergence time for the two major lineages of D. echinata reported herein using the ∼6.4-9.1% COI K2P divergences seen amongst lineages and COI substitution rates reported for a variety of isopods (0.0125 substitutions per site per million years reported by (Ketmaier, Argano & Caccone, 2003) from Stenasellus isopods; 1.56-1.72%divergence per million years reported by Poulakakis & Sfenthourakis (2008) from Orthometopon isopods) suggest D. echinata lineages diverged at least 2.4 million years ago.This suggests that Plio-Pleistocene climactic and oceanographic events (e.g., sea level changes) may have shaped the evolutionary history of D. echinata in the region.Nonetheless, given the poor state of ecological knowledge of South African beach (https://www.sciencedirect.com/topics/agricultural-and-biological-sciences/macrofauna),concerted efforts should be made to better account for the effect of ecological variation in shaping population dynamics of marine species in the region.
the Santamaria laboratory was supported by start-up funds at the University of Tampa and the University of South Florida Sarasota-Manatee.The authors received no external funding for this work.The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Figure 2 SantamariaFigure 3
Figure 2 Phylogenetic patterns of Deto from southern Africa.Results are projected on the majority rule consensus tree produced by analyzing the concatenated mitochondrial dataset of D. echinata in RAxML partitioned by gene under the GTR + model.We observed two monophyletic groups: a ''Namibia-Cape Town'' clade (blues and reds) comprised primarily of individuals collected from Namibia to Bakoven in South Africa, and a ''False Bay-Mossel Bay'' comprising individuals collected from Kommetjie to Glentana in South Africa.Values above branches represent support values for the corresponding branch (top value: Bootstrap Support; bottom: Maximum Posterior Probablities; *: 100 in all analyses).Results of molecular species delimitations are shown as bars, with members of the same putative species clusters identified by color.Bars correspond to: (A) ASAP analyses on the COI dataset, (B) ASAP analyses on concatenated mitochondrial dataset, (C) PTP on tree produced in RAxML assuming a single partition, (D) PTP on tree produced in RAxML assuming partitioning by gene, (E) PTP on tree produced in MrBayes assuming a single partition, (F) PTP on tree produced in MrBayes assuming partitioning by gene, (G) GMYC Relaxed+Yule, (H) GMYC Coalescent, (H) GMYC Strict+Yule.Full-size DOI: 10.7717/peerj.16529/fig-2

Figure 4
Figure 4 Haplotype networks for the nuclear 28S rDNA gene fragment.Colors correspond with those used in other figures.Frequency of haplotype recovery is represented through the relative sizes of the circles.Locality labels correspond with those in all figures and tables.Full-size DOI: 10.7717/peerj.16529/fig-4

Table 2 Estimates of evolutionary divergence between localities as indicated by COI K2P distances.
Minimum and maximum pairwise distances are provided with average distances in parenthesis.The number of base substitutions per site from between sequences are shown.Analyses were conducted using the Kimura 2-parameter model.This analysis involved 71 nucleotide sequences.There were a total of 659 positions in the final dataset.Evolutionary analyses were conducted in MEGA11.