Host and Microbe Scale Processes Jointly Shape Spatial Variation in Aphaenogaster (Hymenoptera: Formicidae)associated Wolbachia

Host and Microbe Scale Processes Jointly Shape Spatial Variation in Aphaenogaster (Hymenoptera: Formicidae)associated Wolbachia | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Host and Microbe Scale Processes Jointly Shape Spatial Variation in Aphaenogaster (Hymenoptera: Formicidae)associated Wolbachia Daniel Malagon, Benjamin Camper, Sophie Millard, Ernesto Recuero, and 6 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8515672/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 13 You are reading this latest preprint version Abstract The spatial distributions of host-associated (HA) microbes are shaped by the spatial processes of environmental selection and dispersal. However, unlike free-living organisms, HA microbes experience selection and dispersal at two separate spatial scales – the scale of the microbes and the scale of their hosts. Therefore, HA microbes must tolerate both the environment created by their host (microbe-scale environment) and the environment in which their host resides (host-scale environment). Likewise, HA microbes can disperse both between hosts through horizontal or vertical transmission (microbe-scale dispersal) and between locations through host movement (host-scale dispersal). In this paper, we examine how host- and microbe-scale spatial processes contribute to the spatial distribution of Wolbachia endosymbionts in Aphaenogaster fulva-rudis-texana (Hymenoptera: Formicidae) complex ants from Great Smoky Mountains National Park. We begin by identifying significant spatial variation in Wolbachia relative abundance at both the host (across the landscape) and microbe (across host lineages) scales. We then demonstrate a correlation between host- and microbe-scale environmental selection, complicating efforts to isolate the independent effects of host- versus microbe-scale processes. To overcome this challenge, we leverage both the broad distributions of individual host lineages across different environments and sites of co-occurrence between different host lineages within the same environments. This allows us to assess how both host- and microbe-scale processes contribute to spatial variation in our system. Ultimately, our results shed light on the myriad of interacting factors governing spatial variation in HA microbes and why spatial variation in HA microbes is more challenging to understand than spatial variation in free-living organisms. Biological sciences/Ecology Earth and environmental sciences/Ecology Biological sciences/Microbiology Aphaenogaster Wolbachia host-associated microbes multiscale environmental selection dispersal limitation Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Two of the most important drivers of a taxon’s spatial distribution are environmental selection and dispersal limitation. 1 Environmental selection (i.e., taxon sorting) causes spatial structure by limiting access of certain taxa to certain locations based on their ability to tolerate local biotic or abiotic conditions. This can yield a wide array of different spatial patterns, including the evolution of ecotypes, genetic isolation by environment (IBE), 2,3 and distance-decay of community similarity. 4 Like environmental selection, dispersal limitation also causes spatial structure by limiting access of certain taxa to certain locations. 5 In the case of dispersal limitation, however, access is limited by the ability of taxa to cross physical barriers and/or unsuitable habitat. 5 Dispersal limitation can result in a wide variety of different spatial patterns, including genetic isolation by distance (IBD), metapopulation dynamics and, once again, distance-decay of community similarity. While the spatial distributions of free-living organisms are primarily governed by environmental selection and dispersal limitation at the scale of the organism, the spatial distributions of host-associated (HA) microbes are governed by environmental selection and dispersal limitation at two scales: the scale of the microbe and the scale of their host. Thus, spatial variation in HA microbes is affected not only by dispersal of microbes from one host to another (microbe scale dispersal limitation) and differences in conditions on the host (microbe scale environmental selection; e.g., due to differences in host genetics or phenotype), but also by dispersal of hosts from one location to another (host scale dispersal limitation) and differences in conditions where the hosts reside (host scale environmental selection). 6 , 7 This makes interpretation of HA microbial spatial patterns challenging. More specifically, because host-scale (i.e., variation in landscape-level environment and host movement) and microbe-scale (i.e., variation in host characteristics, transfer between hosts) processes are often correlated, it can be difficult to disentangle their independent effects on HA microbes. As an example, host genetic distance often correlates with host geographic distance and/or host environmental distance due to processes like genetic drift and local adaptation of the host. 8 – 10 Consequently, the environment external to the host (host-scale) and the environment created by the host (microbe-scale) frequently co-vary, making it difficult to attribute spatial patterns to host- versus microbe-scale environmental selection. Further complicating the interpretation of spatial patterns in HA microbes is the potential for bi-directional interactions across scales. Thus, just as host scale processes can impact the spatial ecology of HA microbes (i.e., top-down effects), microbe scale processes can influence the spatial ecology of their hosts (i.e., bottom-up effects). 11 Many microbes, for instance, alter host dispersal, either as a side-effect of their impacts on host health 12 or because they have evolved to manipulate host dispersal to promote their own transmission. 1314 HA microbes can also alter host environmental selection. In some systems, for example, it is the abiotic tolerances of obligate endosymbionts, rather than the hosts themselves that restrict environmental associations. 15 Beyond dispersal and selection, HA microbes can also manipulate host reproduction. This is important to spatial patterns because it can impact host gene flow between regions. 161718 One of the most notorious examples is Wolbachia 18 , 19 – a widespread 20 bacterium that can affect host genetics by inducing cytoplasmic incompatibility (CI) among hosts harboring differing Wolbachia strains. 21 ,2223, 24 Because of the complicated, correlated and often bi-directional nature of interactions between host- and microbe-scale processes, it can be difficult to determine the degrees to which processes at each scale influence the spatial structure of HA microbes. Regions where cross-scale correlations break down, however, can sometimes be useful to disentangle the effects of host- versus microbe-scale processes. In this paper, we investigate the roles of host- and microbe-scale spatial processes on spatial patterns in Wolbachia carriage among Aphaenogaster fulva-rudis-texana (Hymenoptera: Formicidae; henceforth ‘ A. rudis ’) complex ants in Great Smoky Mountains National Park (GSMNP). Leveraging previously derived, high resolution GSMNP bioclimatic variables, 25,26 we assess the extent to which Wolbachia relative abundance varies as a function of the environment of the host (i.e., host scale environmental selection) and/or spatial distance (i.e., host scale dispersal limitation). We then assess the extent to which Wolbachia relative abundance varies as a function of host genetics (i.e., microbe scale environmental selection) and/or host mitotype (i.e., microbe scale dispersal limitation). Next, we show how host- and microbe-scale processes are strongly correlated, potentially complicating interpretation of spatial patterns. Finally, we repeat our analyses, subsetting our dataset in ways that allow us to break correlations between host- and microbe-scale processes. Ultimately, this allows us to identify the independent roles of processes at both scales as determinants of Wolbachia relative abundance in Aphaenogaster ants from GSMNP. Materials and Methods Study Species Ants of the A. rudis complex are abundant and widespread in eastern hardwood forests of North America. 27 With a generalist strategy that includes colonization of a wide array of forest types, 28 and broad latitudinal, 29 and elevational ranges, 30–34 this complex is ideal for examining effects of host scale environmental selection on HA microbes. Further, A. rudis complex ants within GSMNP comprise four incipient species - A. rudis , A. picea , A. fulva , and A. carolinensis . Generally speaking, A. picea is found at higher elevations while A. rudis is found at lower elevations. 35 Despite these differences, 35,36 members of the complex co-occur extensively, both within GSMNP and across the Southeast broadly, 37 possibly because they are in the early stages of lineage divergence. 37 This makes the A. rudis complex ideal for examining the effects of microbe scale environmental selection (i.e., host genetics) on HA microbes. Study Sites All ants were collected from GSMNP. GSMNP is an ideal location for studying the effects of host scale environmental selection due to its mountainous terrain. In particular, the large elevational gradients characteristic of the GSMNP landscape result in significant bioclimatic variation over relatively small distances. 25 , 26 A. rudis complex ants were collected from 12 All Taxa Biodiversity Index (ATBI) plots 38 as well as an additional plot at one of the lowest elevations in the park (13 total sites) from Fall 2019 - Summer 2021. These plots vary greatly in elevation (522-1,494 meters) and were selected to represent the dominant ecosystems (i.e., distinct environments with different selection pressures) present in GSMNP. At each plot, we collected between 4 and 15 1m x 1m quadrats of litter across 6 to 11 unique time-points, depending on site accessibility. All quadrats within any sampling period were selected to be at least 5m apart to avoid sampling the same colony within multiple quadrats, and at most one ant from any quadrat was selected for sequencing. In addition to quadrat sampling, we spent ~ 30 minutes searching for and aspirating ants during each visit. (For additional details regarding ant collection and identification, see SI I). COI and CAD Gene Extraction, Amplification, and Sequencing Genomic DNA was extracted from 1 leg of each ant using the Thermo Scientific Genejet Genomic DNA purification kit following the manufacturer’s protocol. We amplified the COI gene following Harden et al. 2022 39 using the primers LCO1490 and HCO2198 in 25-µL PCR reactions containing 2.5 µL of template DNA, 1.0 µL of each primer, 2.5 µL of dNTPs, 2.0 µL of MgCl2, 0.125 µL of Platinum Taq, 2.5 µL of buffer, and 13.375 µL of Invitrogen ultra-pure water (Catalog number: 10977023). Reactions were carried out using an Eppendorf MasterCycler with the following settings: initial denaturation stage of 180 seconds at 95 °C followed by 35 cycles of a denaturation stage at 94 °C for 30 seconds, an annealing stage at 45 °C for 30 seconds, an extension stage at 72 °C for 45 seconds, and ending with a final extension at 72 °C for 180 seconds. CAD was amplified using primers CD439F and CD851R. 51 CAD reactions were carried out as above using an Eppendorf MasterCycler beginning with an initial denaturation at 95°C for 3 minutes followed by 40 cycles of a denaturation stage at 94°C for 30 seconds, an annealing stage at 50 °C for 30 seconds, an extension stage at 72 °C for 50 seconds, and a final extension at 72 °C for 5 minutes. Amplified genes were forward and reverse Sanger sequenced at Psomagen. COI and CAD Phylogenetic Tree Construction Forward and reverse COI and CAD Sanger sequencing reads were merged in R (v.4.2.1) 40 using the package sangeranalyseR (v.1.6.1). 41 After quality control, 91 CAD sequences and 80 COI sequences remained from our original 102 samples. We aligned merged reads, without a reference, using our own data as well as data from an A. umphreyi individual that we chose as our outgroup (GenBank Accessions: KP860492.1 and KP730141.1). 42 COI and CAD genes were aligned separately using the MAFFT algorithm in Mesquite. We then employed maximum likelihood methods for phylogenetic tree construction through the web server version (v. 2.4.0) of IQ-Tree 43 using default parameters. This was repeated for the COI and CAD single alignments separately to produce single gene trees. For the COI (mtDNA marker) and CAD (nuDNA marker) genes, the best-fit models of nucleotide substitution, selected based on BIC and using ModelFinder within IQ-Tree, were TPM2u+F+R2 and K2P+R2 respectively. Tree discordance between COI and CAD gene trees was calculated using the cospeciation function and visualized using the cophylo function both in the phytools package in R (version 4.2.1). All statistics were performed using separate COI and CAD trees rather than a concatenated tree. For host-specific analyses, we used all COI and/or CAD sequences available (80 and 91 samples respectively). For analyses that paired COI/CAD and Wolbachia , we only used samples for which both host and Wolbachia data were available (46 samples for COI/ Wolbachia , 52 samples CAD/ Wolbachia , 41 samples COI/CAD/ Wolbachia ). For all analyses, we used a square-root transformation of branch lengths 44 in either the COI or CAD trees. Extraction, Amplification, and Sequencing of Microbial DNA Microbial DNA was extracted using the ZymoBIOMICS DNA Microprep kit (catalogue number D3401). Prior to extraction, individual ants were surface sterilized for 1 minute in 10% bleach solution and rinsed 3 times in ultra-pure water to ensure transient microbes on the exoskeleton were removed. We performed the extraction following the manufacturer’s protocol with the following modifications: 600 µL input of ant lysate was added to 1800 µL of binding buffer due to low input. We skipped filtering steps 4, 12, and 13 due to low input and heated the elution buffer to 60 °C prior to elution to maximize DNA concentrations. Samples were lysed for 40 minutes using a vortex genie and for 20 minutes using a tissuelyzer2 (Qiagen) at 25 x/second to effectively lyse the ant exoskeleton. We characterized entire bacterial communities by amplifying the V4 region of the 16S rRNA gene using touchdown PCR with common dual index primers. 45 Briefly, we amplified in 20-µL PCR reactions containing 5 µL DNA, 5 µL of the primer set (515F and 806R), 2 µL of 10x Accuprime PCR Buffer II, 0.15 µL of Accuprime HiFi polymerase, and 7.85 µL of ultra-pure water per reaction. Reactions were carried out using an Eppendorf MasterCycler with the following settings: an initial denaturation stage of 120 seconds at 95 °C followed by 20 cycles of a denaturation stage at 95 °C for 20 seconds, an annealing stage at 60 °C for 20 seconds, and an extension stage at 72 °C for 5 minutes. This was followed by 20 cycles of a denaturation stage at 95 °C for 20 seconds, an annealing stage at 55 °C for 20 seconds, an extension stage at 72 °C for 5 minutes, and a final extension stage at 72 °C for 10 minutes. Samples were normalized prior to pooling using a SequalPrep Normalization Plate Kit (catalogue number #A1051001). Pooled sample library concentrations were quantified via KAPA qPCR (Roche, cat. #07960140001). Multiplexed pooled libraries were sequenced paired end 2 x 300 cycles on an Illumina NextSeq 2000 to an average depth of 245,448 reads. Resulting FASTQ files were analyzed using a custom Qiime2 46 pipeline. (For additional information on Taxonomic Assignment and Dataset Manipulation, see SI I; see also Github repository below). Environmental axis We considered elevation, mean soil moisture (volumetric content), soil minimum temperature, and soil maximum temperature derived from previous GSMNP studies. 25,26 We first z-transformed these four environmental variables, and then calculated environmental distances based on the Euclidean distances between the z-transformed environmental variable values at each capture location. Next, we performed a principal component analysis (PCA) based on the environmental variables associated with each ant. We then selected the first principal component (PC1; i.e., the axis capturing the most variation in the system) and used the value of each ant along this axis for subsequent analysis of environmental effects. Statistical Analyses All statistical analyses were performed in R (v.4.2.1).All data and code can be found in the following github repository: https://github.com/dmalago/Aphaenogaster_GSMNP_Microbiome. Spatial autocorrelation : We used the jitter function from base R tojitter site values for latitude and longitude so that each ant sample had a unique location. We then used the chooseCN function from the adespatial package to create a connection network. This was done using Delaunay triangulation (type 1). From this network, we used the nbdists function from the spdep package to calculate edge distances. Next, we used the nb2listw function from the spdep package to find row-standardized (style = 4) spatial weights for network edges. Finally, we inputted Wolbachia relative abundances and the list of spatial weights into the moran.test function, also from the spdep package, to determine whether there was significant autocorrelation. Results were visualized using the moran.plot function, again from the spdep package. Regression Analyses : We used generalized linear models 47 with beta distributions to regress Wolbachia relative abundances against principal component one (PC1) of the environmental PCA ( Wolbachia ~ Environment , see above for axis definition). This was done for all ants and for ants from individual lineages separately. We also regressed Wolbachia relative abundance against host COI and CAD clade/group ( Wolbachia ~ COI Clade *CAD Clade , see Results for information on clades/groups). All regressions were performed using the betareg function from the betareg package. Heatmaps : We constructed heatmaps of Wolbachia relative abundance for each Wolbachia ASV in each ant using the pheatmap function from the pheatmap package. This was done ordering ants according to both the COI and the CAD phylogenies. Mantel tests : We used Mantel 48 tests to assess whether host relatedness varied with geographic distance and whether host relatedness varied with environmental similarity. All Mantel tests were performed using the mantel function from the vegan package(v. 2.6.4). 49 We then used the mantel.correlog function, also from the vegan package (v. 2.6.4), to calculate mantel correlations at varying geographic or environmental distance classes. For spatial distance, we used the Haversine formula to calculate the great circle distance between all ant capture locations based on jittered (see above) longitude and latitude. This was implemented using the distm function from the geosphere package (v. 1.5-20). For environmental distance, we considered the distance between ants along our PC1 environmental axis (see Environmental axis , above). For host relatedness, we used distances based on the branch lengths of the square root transformed COI or CAD trees. Branch lengths were calculated using the cophenetic.phylo function from the ape package (v. 5.7.1). 50 Paired sample Wilcoxon tests: We used paired sample Wilcoxon tests 51 to test for differences in Wolbachia relative abundance between host clades/groups (see Results ), controlling for environment/spatial autocorrelation. First, we identified sites where different COI clades/groups were co-localized. We then averaged Wolbachia relative abundances for all ants from each clade/group at each site of co-localization. This gave a single Wolbachia relative abundance for each clade/group at each site. We then used these average Wolbachia relative abundances in paired sample Wilcoxon tests, pairing ants from different clades/groups at the same sites. Results Ant Clades/Groups Using the COI consensus tree (Fig. S2.1 – left side), we visually identified 2 main clades (henceforth the “blue” and “purple” clades) as well as a set of basal ants (henceforth the “red” [paraphyletic] group, see also SI Figure S3.6B for haplotype maps suggesting these are a separate group but not necessarily basal to the blue and purple clades). Using the CAD consensus tree (Fig. S2.1 – right side), we visually identified two main clades (henceforth the “green” and “yellow” clades) as well as several basal ants (henceforth the “black” group). Although consensus trees for COI and CAD sequences displayed significantly different topologies (RF distance = 148, p = 1), there was a relatively high degree of congruence between the red, purple and blue COI clades/groups and the green and yellow CAD clades. In particular, ants with red group COI sequences were significantly more likely to have green clade CAD sequences, while ants with blue clade COI sequences were significantly more likely to have yellow clade CAD sequences (Fisher’s exact test: p = 0.00067, see SI II Figure S2.2, Table S2.1). Ant Microbiota Based on regression analyses (see Statistical Analyses ), Wolbachia was the only microbial taxon whose relative abundance varied significantly across either environment or host clade (see SI III Figures S3.1 and S3.2). Consequently, we chose to focus on elucidating spatial patterns in Wolbachia carriage. Thirty-two separate ASVs were taxonomically classified to the Wolbachia genus, including members of supergroups A, B and F (see SI III Figure S3.3). 18 Only three, however, independently comprised more than 0.01% of reads across all ant microbiota (two from supergroup A, and one from supergroup F). These were also the only Wolbachia ASVs found in more than one ant. Host-scale variation in Wolbachia Wolbachia relative abundances exhibited significant spatial variation across GSMNP (see Figure 1). Whereas Wolbachia was present on most ants from the western side of the park, Wolbachia was absent or present in very low abundances (<0.05%, see Figure 1A) on most ants from the eastern side of the park. This resulted in significant positive spatial autocorrelation in Wolbachia relative abundance (see Figure 1B). Regressing Wolbachia relative abundance against the PC1 environmental axis further indicated that Wolbachia relative abundance was higher in ants from warmer, drier, lower elevation regions of the park and that Wolbachia relative abundance was lower in ants from cooler, wetter, higher elevation regions of the park (see Figure 1C). This was consistent with overall spatial trends, since the western side of the park is generally warmer, drier and lower elevation than the eastern side of the park. Using variance partitioning to separate the spatial versus environmental drivers of Wolbachia abundance, we found that both spatial distance (9% of variation) and the interaction between spatial distance and environment (16% of variation) were important for explaining differences in Wolbachia relative abundances across ants (see SI III Figure S3.5). Microbe-scale variation in Wolbachia Wolbachia relative abundance varied significantly with both mitochondrial and nuclear sequence variation of the host (see Figure 2). More specifically, our regression model suggested that ants with red group and purple clade COI sequences had significantly higher Wolbachia relative abundances as compared to ants with blue clade COI sequences, and that ants with red group and (marginally) purple clade COI sequences had significantly lower Wolbachia relative abundances when they had yellow as compared to green clade CAD sequences. Closer inspection suggested that three distinct Wolbachia ASVs were responsible for the majority of Wolbachia in our system (see Figure 3, see also SI III figure S3.6). Two of the three dominant ASVs (henceforth wArudA1, wArudA2) were closely related, clustering together insupergroup A and differing from each other by 1/214 base pairs (bp). By contrast, the third dominant ASV (henceforth wArudF1) was more distantly related, clustering in supergroup F and differing from wArudA1 and wArudA2 by 8/214 and 9/214 bp respectively (see SI III, Figure S3.3). Notably, these three ASVs showed strong variation across COI ant clades, indicating a likely link between mitotype and the presence of specific Wolbachia variants (see Figure 3A, see also SI III Figures S3.6B and S3.7). In particular, whereas ants from the red clade had all three Wolbachia ASVs, ants from the purple clade had only wArudA2, and ants from the blue clade had none of the three Wolbachia ASVs. As compared to host COI sequences, the presence/absence of individual Wolbachia ASVs was not as strongly correlated with host CAD sequences. Nevertheless, consistent with the lower abundance of Wolbachia in yellow clade ants overall, a larger proportion of green clade ants had all three Wolbachia ASVs, while a larger proportion of yellow clade ants had no Wolbachia ASVs (see Figure 3B, see also SI III Figures S3.6C and S3.8). Correlation between host- and microbe-scale drivers of variation Mantel tests based on Spearman, but not Pearson correlation coefficients demonstrated a significant relationship between COI sequence similarity and geographic distance (Spearman: r: 0.1005, p = 0.020; Pearson: r = 0.0703; p = 0.082; Figure 4A, see also SI II Figure S2.3A,B). Likewise, Mantel tests based on both Spearman and Pearson correlation coefficients demonstrated a significant relationship between COI sequence similarity and environmental similarity (Spearman: r = 0.0912, p = 0.014; Pearson: r = 0.0685; p = 0.042; Figure 4C, see also SI II Figure S2.3C,D). Distance-based redundancy analysis (dbRDA) and variance partitioning supported results from individual Mantel tests, suggesting that both distance and distance + environment explained the similarity of COI sequences (see SI II Figure S2.5A). Consistent with the observed correlations between COI sequence similarity and both spatial and environmental distance, ants with red group, purple clade and blue clade COI sequences exhibited different distributions across GSMNP. Red group ants dominated lower elevations on the western side of the park, while blue clade ants dominated higher elevations on the eastern side of the park. Purple clade ants were predominantly found towards the center of the park where they co-occurred with the other two clades/groups across a range of elevations (see Figure 5A). In accordance with their differing distributions, environmental conditions were also significantly different across the three COI clades/groups (see Figure 5B; Kruskal-Wallis chi-squared = 26.681, df = 2, p-value = 1.6E-6, see also SI II Figure S2.6A). Ants from the red group were associated with warmer, drier, and lower elevation conditions while ants from the blue clade were associated with colder, moister and higher elevation conditions. Ants from the purple clade were intermediate, though differences between the blue and the purple clades were not significant. Unlike COI sequences, no significant relationships were observed between CAD sequence similarity and either geographic distance (Spearman: r: -0.0902, p = 0.978; Pearson: r = -0.0634 ; p = 0.882; Figure 4B, see also SI II Figure S2.4A,B) or environmental distance (Spearman: r = -0.0319, p =0.814 ; Pearson: r = -0.0379, p = 0.832; Figure 4D, see also SI II Figure S2.4C,D). Likewise, CAD sequences did not exhibit obviously different distributions across GSMNP (see Figure 5C), nor were environmental conditions significantly different between the green and yellow clades (see Figure 5D, Kruskal-Wallis chi-squared = 2.0236, df =1, p-value = 0.1549; see also SI II Figure S2.6B). Isolating host-scale drivers of variation in Wolbachia We examined the role of environment (host-scale process) while controlling for COI sequence (microbe-scale process) by considering the effects of the PC1 environmental variable on Wolbachia carriage by blue clade and red group ants separately. When we did this, we found that overall Wolbachia abundance, as well as abundance of each of the three dominant Wolbachia ASVs (wArudA1, wArudA2, and wArudF1) varied in response to environment (see Figure 6A,C-E) for red group ants. In particular, there was a precipitous decline in Wolbachia at a threshold environmental PC1 value of approximately -2 (roughly corresponding to a minimum soil temperature of -5 ° C, a soil moisture of 0.073 % volumetric moisture content, a soil temperature range of 32.5 ° C and an elevation of 622 m) with the two ants from the two coldest sites (Oconaluftee (OCO), Trillium Gap (TG)) exhibiting much lower (<0.5%) Wolbachia relative abundance as compared to ants from other sites. While the ant from Trillium Gap was yellow clade for its CAD sequence (i.e., potentially lower Wolbachia due to host genetics), the ant from Oconaluftee had a basal (‘black group’) CAD sequence. Unlike red group ants, the relationship between Wolbachia and environment was less apparent for blue clade ants, only reaching significance for overall Wolbachia relative abundance and the abundance of the wArudA2 ASV and, even then, only reaching significance when an outlier ant from Ramsey Cascades was removed. Notably, this was a generally unusual ant containing both wArudA2 and wArudF1 (most other ants either contained all three Wolbachia ASVs or else only wArudA2, see SI III Figure S3.6). The weaker relationship between Wolbachia and environment in blue clade ants is not surprising given the dramatically lower Wolbachia relative abundances in blue clade ants in general. Overall, within-clade comparisons, particularly for red group ants, suggest that host-scale processes (environmental conditions and/or host dispersal limitation) independently impact the spatial distribution of Aphaenogaster -associated Wolbachia across GSMNP. Isolating microbe-scale drivers of variation in Wolbachia We examined the effects of COI sequence (microbe-scale process) while controlling for environment (host-scale process) by contrasting average Wolbachia carriage of ants from different COI clades/groups paired according to collection site. When we did this, we found that red group and blue clade ants co-occurred at four sites, and that the difference in Wolbachia carriage was borderline in significance. Likewise, we found that red group and purple clade ants and purple clade and blue clade ants co-occurred at two and three sites respectively, and that the differences in Wolbachia carriage were not significant. Finally, we found that red group + purple clade and blue clade ants co-occurred at seven sites, and that the difference in Wolbachia carriage was highly significant (see Table 1). In addition, we examined the effects of CAD sequence (microbe-scale process) while controlling for environment (host-scale process) by contrasting Wolbachia carriage across ants from different CAD clades paired according to collection site. Again, when we did this, we found that green and yellow clade ants co-occurred at nine sites, and that the difference in Wolbachia carriage was significant. Finally, we examined the effects of COI/CAD sequence while controlling for environment. In this case, we found that red + purple/yellow clade ants and red + purple/green clade ants co-occurred at four sites, and the difference in Wolbachia carriage was not significant. Overall, these comparisons across paired samples suggest that microbe-scale processes (host genetics and/or host-to-host transmission), particularly processes correlated with COI sequence, independently impact the spatial distribution of Aphaenogaster -associated Wolbachia across GSMNP. Discussion In this paper, we examined host- and microbe-scale patterns in Wolbachia relative abundance among A. rudis complex ants from across Great Smoky Mountains National Park (GSMNP). Our goal was to determine whether spatial structure was primarily attributable to host-scale processes, microbe-scale processes or a combination of the two (i.e., multiscale processes). Independent effects of host-scale processes We inferred independent effects of host-scale processes – more specifically, host-scale environmental selection – based on how environment impacted Wolbachia load across individual ant COI clades/groups. In particular, we found that some combination of low temperature, high moisture and high elevation lowers Wolbachia abundance/persistence in our system (see below). This is not unprecedented. 52,53 In a recent study of captive A. rudis colonies, for example, Wolbachia relative abundance decreased in response to artificial warming. 54 While this trend is opposite what we observed, it nonetheless indicates that temperature can affect Wolbachia abundance. Indeed, differences in directionality could merely reflect the existence of an optimum in the Wolbachia thermal performance curve. Notably, ants in the captive study were warmed to a constant temperature of 32°C, which is much higher than the ambient annual temperatures experienced by even the lowest elevation ants in our system (though not necessarily higher than the maximum annual temperature at the warmest sites). Thus, our range of temperatures may have been below the optimal growth temperature for Aphaenogaster -associated Wolbachia , causing an increase in Wolbachia with temperature. By contrast, the captive study may have heated ants past the optimal growth temperature, causing a decrease in Wolbachia with temperature. This is in keeping with studies of other Hymenoptera species 55 which have demonstrated reduced Wolbachia carriage (e.g., due to temperature dependent interactions with a Wolbachia phage) at both excessively low and excessively high temperatures. Interestingly, a recent meta-analysis 56 across ~2500 species of arthropods also hints at the possibility of a unimodal response of Wolbachia to temperature, with Wolbachia prevalence increasing with temperature in temperate zones (i.e., from a lower baseline temperature) and decreasing with temperature in tropical zones (i.e., from a higher baseline temperature). While the host-scale patterns that we observed are well explained by host-scale environmental selection, they are not as easily explained by host-scale dispersal limitation. If host-scale dispersal limitation was fully responsible for the spatial distribution of Aphaenogaster -associated Wolbachia in GSMNP, then, in keeping with our findings, there should be fewer Wolbachia -positive ants on the eastern side of the park. However, Wolbachia -positiveants that do reach the eastern side of GSMNP should have similar Wolbachia loads to their Wolbachia -positive counterparts on the western side of the park(i.e., the frequency of Wolbachia -positive ants should decrease, but the relative abundance of Wolbachia on Wolbachia- positiveants should be similar everywhere). The fact that both prevalence and load decreased along the West-to-East gradient indicates that host dispersal is not the sole driver of Wolbachia spatial patterns. That said, host-scale dispersal limitation does seem to exist, at least based on COI sequence (see SI II Fig. S2.5A) and thus, could partially explain the smaller fraction of Wolbachia -positive ants on the East side of the park - even if it doesn’t explain the overall lower Wolbachia loads in eastern ants when Wolbachia is present. Independent effects of microbe-scale processes We inferred independent effects of microbe-scale processes based on the significantly different Wolbachia loads across host mitotypes, even in paired samples from the same sets of sites. Notably, this pattern could have emerged either through microbe-scale environmental selection or microbe-scale dispersal limitation. Microbe-scale environmental selection would indicate that the mitochondrial COI sequence is correlated (not necessarily causally) with some aspect of host biology (e.g., immune system, behavior, food preference) that impacts Wolbachia performance. Again, this would not be unprecedented. Among leaf-cutter ants ( Acromyrmex echinatior ), for example, different patrilines exhibit significant differences in Wolbachia densities. 57 Meanwhile among Vollenhovia emeryi , genotypes associated with different wing morphology exhibit different Wolbachia colonization rates. 58 Even among Aphaenogaster , a recent study identified differences in Wolbachia abundances between A. rudis and A. fulva ants, although these results were based on only three A. fulva and two A. rudis colonies. 59 The alternate explanation – that Wolbachia are limited to red group ants because of microbe-scale dispersal limitation – would indicate a barrier to Wolbachia transmission among different host mitotypes. Again, this is consistent with what we know about Wolbachia biology.Broadly speaking, Wolbachia has two modes of microbe-scale dispersal: horizontal transmission and vertical transmission. Although extensive evidence exists suggesting that Wolbachia can be acquired horizontally across large host phylogenetic distances, 60 including among ants, 61 horizontal transmission within species requires a number of relatively restrictive conditions. 62 This likely limits the extent to which Wolbachia can transfer directly from one ant to another (microbe-scale dispersal limitation), particularly over shorter, ecological timescales. Consequently, in our system, microbe-scale dispersal of Wolbachia likely occurs predominantly via vertical transmission (mother to offspring). 18 As a result, Wolbachia colonization should closely mirror patterns in mtDNA inheritance, even if there is no correlation between mitochondrial COI sequence and other aspects of host biology impacting Wolbachia fitness (i.e., no microbe-scale environmental selection). Unfortunately, in our system, it is difficult to determine whether microbe-scale environmental selection, microbe-scale dispersal limitation or a combination of both underlies differences Wolbachia carriage across COI sequences. Nevertheless, the extremely tight coupling between COI sequence and each of the three Wolbachia ASVs, along with the limited number of Wolbachia acquisitions and losses predicted by ancestral state reconstructions (ASRs, see SI III, Figures S.3.7 and S3.8), suggests that dispersal is strongly limited to vertical transmission. This makes microbe-scale dispersal limitation a strong contender for explaining the restriction of Wolbachia to red group ants and, to a lesser extent, purple clade ants. Consistent with this hypothesis, we did not find any relationships between other bacterial taxa (including both other endosymbionts and ectosymbionts), and COI sequence (see SI III, Figures S3.1 for visualizations), although it is possible that some of the other putative endosymbionts (e.g., Sulcia )were only detected because they were present in prey from Aphaenogaster stomach contents. Like COI sequence, CAD sequence also appears to be correlated with some aspect of host biology that impacts Wolbachia performance. Notably, we found an effect of CAD sequence even within COI clades, suggesting that the correlation between CAD sequence and Wolbachia is not fully explained by the same drivers that underlie the correlation between COI sequence and Wolbachia . Likewise, CAD sequence remained correlated with Wolbachia load even when controlling for environment, as would be expected given that CAD is not, itself, correlated with environment. Unfortunately, the relatively few western locations harboring red clade ants, coupled with the low capture rate of red clade ants at eastern locations, made it difficult for us to test the effect of CAD within COI clades while simultaneously controlling for environment. Rather, we had only four sites for our paired Wilcoxon test, which did not, as a result, rise to the level of significance. Nevertheless, there was still evidence of a weak trend, suggesting that, with additional statistical power, it might be possible to detect independent effects of COI and CAD independent of environment (see Table 1). Regardless of whether COI and CAD sequences reflect the same or different aspects of host biology underlying Wolbachia load, there is no question that microbe-scale processes are an independent contributor to the spatial distribution of Aphaenogaster -associated Wolbachia across GSMNP. Possible interactions between host and microbe scales While we found conclusive evidence for independent effects of both host- and microbe-scale processes driving patterns in Aphaenogaster- associated Wolbachia abundance across GSMNP, we also identified ample opportunity for multiscale interactions. Admittedly, such interactions are more difficult to demonstrate. Nevertheless, it is worth speculating on potential mechanisms that might connect host- and microbe-scale processes. Based on recent findings in Drosophila , for example, it is possible that colder host environments (i.e., host-scale environmental selection) reduce vertical transmission (i.e., microbe-scale dispersal limitation) of Wolbachia to Aphaenogaster alates. 63 This could explain the lower prevalence (though not necessarily the lower abundance) of Wolbachia in ants, including red group ants, from the colder Eastern regions of the park. Another possible multiscale interaction is reproductive manipulation. If Wolbachia has been prevalent on the western side of the park for a long time, and if Wolbachia induces CI or some other form of reproductive manipulation in A. rudis ants, then this could prevent western invasion of blue clade COI, even as blue clade nuclear genes (i.e., yellow clade CAD) become incorporated into red group populations (i.e., the microbe alters host-scale gene flow). Consistent with this hypothesis, we observed a higher percentage of red group COI (30%) mismatched with yellow clade CAD as compared to blue clade COI (21%) mismatched with green clade CAD. This is what would be expected if the offspring of uninfected females and infected males are non-viable but the offspring of infected females and uninfected males are viable. Further, while mismatch in blue clade COI is unexpected, this may be explained by mismatch that existed prior to Wolbachia colonization (akin to incomplete lineage sorting) or as a result of incomplete CI. Notably, temperature differences can result in differing levels of both CI penetrance and Wolbachia density, 55,64,65 providing a mechanism by which CI might serve as a stronger barrier to gene flow on the western side of the park, but a weaker barrier on the eastern side of the park. Another potential multiscale mechanism that may contribute to our findings is an effect of Wolbachia on ant dispersal. This could occur either because Wolbachia alters the movement patterns of alates or because it reduces production of alates entirely. 66 In both cases, the effect would be to slow the eastern spread of red group COI, even if CI gives the red group a long-term competitive advantage. 67 Finally, Wolbachia could prevent red group ants from penetrating eastward by lowering survival in the colder, wetter, higher elevation regions of the park. Again, this multiscale interaction could restrict the spatial extent of the red group COI, even if it gains a competitive demographic advantage through CI. Conclusions The complex genetic structure of A. rudis ants, their propensity to inhabit a wide range of different environments across large spatial extents, and their colonization by multiple strains of the notoriously complex Wolbachia endosymbiont makes A. rudis an ideal system for studying the interplay between environmental selection and dispersal limitation at the host and microbe scales. Our findings suggest that this interplay leads to interesting spatial patterns across the mountainous terrain in GSMNP. Further, multiple opportunities exist for extending the current study. This includes better characterizing Wolbachia differences in the contact zone between COI clades/groups, undertaking more extensive comparisons of Wolbachia carriage in ants with and without COI/CAD mismatch and extending similar analyses from GSMNP to the Southern Appalachians more broadly in order to better tease apart the effects of environment versus distance in shaping patterns of Wolbachia prevalence and abundance. Beyond field characterization, another fruitful avenue would be lab studies examining how Wolbachia impacts host thermal tolerance, how temperature effects Wolbachia load and transmission, or even whether and how Wolbachia induces CI or other forms of reproductive manipulation. As keystone mutualists, Aphaenogaster ants have garnered significant attention across various research areas, 27,29,33,42,68–72 including recent explorations into their microbial associations. 59,73,74 Our study suggests that, even beyond their importance to processes like seed dispersal and pest control, 75 A. rudis complex antsmight serve as model systems for understanding the multiscale, bidirectional interactions between hosts and their HA microbes. Declarations Funding Declaration : This research was made possible, in part, with support from the Clemson University Genomics and Bioinformatics Facility, which receives support from the College of Science and two Institutional Development Awards (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health under grant numbers P20GM146584 and P20GM139769. Author Contribution D.A.M., B.C. and S.B. designed the experiment and performed the fieldwork. D.A.M., C.H., E.R.G. and M.C. designed and performed sequencing of the host. D.A.M., B.C., S.B., M.A.S., S.M. and A.S. designed and performed sequencing of the microbiota. D.A.M., S.A.B., S.R.B. and S.R.B. designed and performed Wolbachia analysis. All authors contributed to data analysis and writing of the manuscript. Acknowledgement We thank Dr. Megan Pitz for her help with creation of Figure 3 and her willingness to tolerate centipedes in her closet. We would also like to thank Nico Malagon, Dr. Tatsuhiro Kato, Daniel Cross, Jackson Lloyd, Drew Kanes, Simon Dunn, and Mr. Paul Super for their help in the collection of ant samples. Data Accessibility and Benefit-Sharing: Data Accessibility :The datasets generated and/or analysed during the current study are available in the National Center for Biotechnology Information (NCBI) sequence read archive (SRA) under project number PRJNA1347072. All data and code (BIOM files, metadata files, and all R code) necessary for the analyses presented in this manuscript are additionally available at https://github.com/dmalago/Apheanogaster_GSMNP_Microbiome. Benefit-Sharing : Research results have been shared with Great Smoky Mountains National Park through yearly reports. References Bellier, E. et al. Distance decay of similarity, effects of environmental noise and ecological heterogeneity among species in the spatio-temporal dynamics of a dispersal‐limited community. Ecography 37 , 172–182 (2014). Glück, M., Geue, J. C. & Thomassen, H. A. Environmental differences explain subtle yet detectable genetic structure in a widespread pollinator. BMC Ecol. Evol. 22 , 8 (2022). McRae, B. H. ISOLATION BY RESISTANCE. Evol. (N Y) . 60 , 1551–1561 (2006). Soininen, J., McDonald, R. & Hillebrand, H. The distance decay of similarity in ecological communities. Ecography 30 , 3–12 (2007). Baguette, M., Blanchet, S., Legrand, D., Stevens, V. M. & Turlure, C. Individual dispersal, landscape connectivity and ecological networks. Biol. Rev. 88 , 310–326 (2013). Couch, C. E., Epps, C. W. & Host Microbiome, and Complex Space: Applying Population and Landscape Genetic Approaches to Gut Microbiome Research in Wild Populations. J. Hered. 113 , 221–234 (2022). Ruuskanen, M. O., Sommeria-Klein, G., Havulinna, A. S., Niiranen, T. J. & Lahti, L. Modelling spatial patterns in host‐associated microbial communities. Environ. Microbiol. 23 , 2374–2388 (2021). Orsini, L., Vanoverbeke, J., Swillen, I., Mergeay, J. & De Meester, L. Drivers of population genetic differentiation in the wild: isolation by dispersal limitation, isolation by adaptation and isolation by colonization. Mol. Ecol. 22 , 5983–5999 (2013). Hutchison, D. W., Templeton, A. R., CORRELATION OF PAIRWISE GENETIC & AND GEOGRAPHIC DISTANCE MEASURES: INFERRING THE RELATIVE INFLUENCES OF GENE FLOW AND DRIFT ON THE DISTRIBUTION OF GENETIC VARIABILITY. Evol. (N Y) 53 , 1898–1914 (1999). Wright, S. ISOLATION DISTANCE Genetics 28 , 114–138 (1943). Li, G. et al. Host-microbiota interaction helps to explain the bottom-up effects of climate change on a small rodent species. ISME J. 14 , 1795–1808 (2020). Risely, A., Klaassen, M. & Hoye, B. J. Migratory animals feel the cost of getting sick: A meta-analysis across species. J. Anim. Ecol. 87 , 301–314 (2018). Andriolli, F. S. et al. Do zombie ant fungi turn their hosts into light seekers? Behav. Ecol. 30 , 609–616 (2019). Vyas, A., Kim, S. K., Giacomini, N., Boothroyd, J. C. & Sapolsky, R. M. Behavioral changes induced by Toxoplasma infection of rodents are highly specific to aversion of cat odors. Proceedings of the National Academy of Sciences 104, 6442–6447 (2007). Lemoine, M. M., Engl, T. & Kaltenpoth, M. Microbial symbionts expanding or constraining abiotic niche space in insects. Curr. Opin. Insect Sci. 39 , 14–20 (2020). Otti, O. Genitalia-associated microbes in insects. Insect Sci. 22 , 325–339 (2015). Telschow, A., Hammerstein, P. & Werren, J. H. The Effect of Wolbachia on Genetic Divergence between Populations: Models with Two-Way Migration. Am. Nat. 160 , S54–S66 (2002). Kaur, R. et al. Living in the endosymbiotic world of Wolbachia: A centennial review. Cell. Host Microbe . 29 , 879–893 (2021). Weeks, A. R., Reynolds, T., Hoffmann, A. A. & K. & Wolbachia dynamics and host effects: what has (and has not) been demonstrated? Trends Ecol. Evol. 17 , 257–262 (2002). Weinert, L. A., Araujo-Jnr, E. V., Ahmed, M. Z. & Welch, J. J. The incidence of bacterial endosymbionts in terrestrial arthropods. Proceedings of the Royal Society B: Biological Sciences 282, 20150249 (2015). Calvitti, M., Marini, F., Desiderio, A., Puggioli, A. & Moretti, R. Wolbachia Density and Cytoplasmic Incompatibility in Aedes albopictus: Concerns with Using Artificial Wolbachia Infection as a Vector Suppression Tool. PLoS One . 10 , e0121813 (2015). Tortosa, P. et al. Wolbachia Age-Sex-Specific Density in Aedes albopictus: A Host Evolutionary Response to Cytoplasmic Incompatibility? PLoS One . 5 , e9700 (2010). Jaenike, J., Dyer, K. A., Cornish, C. & Minhas, M. S. Asymmetrical Reinforcement and Wolbachia Infection in Drosophila. PLoS Biol. 4 , e325 (2006). Cruz, M. A., Magalhães, S., Sucena, É. & Zélé, F. Wolbachia and host intrinsic reproductive barriers contribute additively to postmating isolation in spider mites. Evol. (N Y) . 75 , 2085–2101 (2021). Fridley, J. D. Downscaling Climate over Complex Terrain: High Finescale (<1000 m) Spatial Variation of Near-Ground Temperatures in a Montane Forested Landscape (Great Smoky Mountains)*. J Appl Meteorol Climatol 48, 1033–1049 (2009). Stark, J. R. & Fridley, J. D. Topographic Drivers of Soil Moisture Across a Large Sensor Network in the Southern Appalachian Mountains (USA). Water Resour. Res 59 , (2023). Lubertazzi, D. The biology and natural history of Aphaenogaster rudis. Psyche (Stuttg) . 10.1155/2012/752815 (2012). Paysen, E. E. S. & DIVERSITY AND ABUNDANCE OF ANTS AT FOREST EDGES IN THE GREAT SMOKY MOUNTAINS NATIONAL PARK. undefined-undefined (2007). Helms Cahan, S. et al. Modulation of the heat shock response is associated with acclimation to novel temperatures but not adaptation to climatic variation in the ants Aphaenogaster picea and A. rudis. Comp. Biochem. Physiol. Mol. Integr. Physiol. 204 , 113–120 (2017). Lessard, J. P., Dunn, R. R., Parker, C. R. & Sanders, N. J. Rarity and Diversity in Forest Ant Assemblages of Great Smoky Mountains National Park. Southeast. Nat. 6 , 215–228 (2007). Lynch, J. F. & Seasonal Successional, and Vertical Segregation in a Maryland Ant Community. Oikos 37 , 183 (1981). Fellers, J. H. Interference and exploitation in a guild of woodland ants. Ecology 68 , 1466–1478 (1987). Banschbach, V. S., Brunelle, A., Bartlett, K. M., Grivetti, J. Y. & Yeamans, R. L. Tool use by the forest ant Aphaenogaster rudis: Ecology and task allocation. Insectes Soc. 53 , 463–471 (2006). Cole, A. C. A Guide to the Ants of the Great Smoky Mountains National Park, Tennessee. Am. Midl. Nat. 24 , 1–88 (1940). Warren, R. J. et al. Climate-driven range shift prompts species replacement. Insectes Soc. 63 , 593–601 (2016). Warren, R. J., McAfee, P. & Bahn, V. Ecological differentiation among key plant mutualists from a cryptic ant guild. Insectes Soc. 58 , 505–512 (2011). Buono, C. M. et al. Uncovering how behavioral variation underlying mutualist partner quality is partitioned within a species complex of keystone seed-dispersing ants. Insectes Soc. 69 , 247–260 (2022). Nichols, B. J. & Langdon, K. R. The Smokies All Taxa Biodiversity Inventory: History and Progress. Southeastern Naturalist 6 , (2007). Caterino, M. S. & Harden, C. W. Unseeing and Unseen: On the Distribution, Morphology, and Larva of One of North America’s Rarest Histerid Beetles, Geocolus caecus Wenzel (Coleoptera: Histeridae). Coleopt Bull 76 , (2022). Team, R. C. & R Core Team. R Core Team. & R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL (2014). http://www.R-project.org/. R Foundation for Statistical Computing (2014). Chao, K. H., Barton, K., Palmer, S. & Lanfear, R. sangeranalyseR: Simple and Interactive Processing of Sanger Sequencing Data in R. Genome Biol. Evol 13 , (2021). DeMarco, B. B. & Cognato, A. I. A multiple-gene phylogeny reveals polyphyly among eastern North American Aphaenogaster species (Hymenoptera: Formicidae). Zool. Scr. 45 , 512–520 (2016). Trifinopoulos, J., Nguyen, L. T., von Haeseler, A. & Minh, B. Q. W-IQ-TREE: a fast online phylogenetic tool for maximum likelihood analysis. Nucleic Acids Res. 44 , W232–W235 (2016). Letten, A. D. & Cornwell, W. K. Trees, branches and (square) roots: why evolutionary relatedness is not linearly related to functional distance. Methods Ecol. Evol. 6 , 439–444 (2015). Kozich, J. J., Westcott, S. L., Baxter, N. T., Highlander, S. K. & Schloss, P. D. Development of a Dual-Index Sequencing Strategy and Curation Pipeline for Analyzing Amplicon Sequence Data on the MiSeq Illumina Sequencing Platform. Appl. Environ. Microbiol. 79 , 5112–5120 (2013). Bolyen, E. et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat. Biotechnol. 37 , 852–857 (2019). Nelder, J. A. & Wedderburn, R. W. M. Generalized Linear Models. J. R Stat. Soc. Ser. A . 135 , 370 (1972). Mantel, N. The detection of disease clustering and a generalized regression approach. Cancer Res (1967). Jari Oksanen, R. et al. Package ‘vegan’. (2007). https://www.researchgate.net/publication/228339454_The_vegan_Package?enrichId=rgreq-87dd17328dd562faca4cd6cd67fd8d2b-XXX&enrichSource=Y292ZXJQYWdlOzIyODMzOTQ1NDtBUzoxMDIxMjA 3NjYzMTI0NjVAMTQwMTM1ODg5NzY3NQ%3D%3D⪙=1_x_2&_esc=publicationCoverPdf http://cran.ism.ac.jp/web/packages/vegan/vegan.pdf Paradis, E. & Schliep, K. ape 5.0: an environment for modern phylogenetics and evolutionary analyses in {R}. Bioinformatics 35 , 526–528 (2019). Conover, W. J. Practical Nonparametric Statistics . (1999). MOUTON, L., HENRI, H., BOULETREAU, M. & VAVRE, F. Effect of temperature on Wolbachia density and impact on cytoplasmic incompatibility. Parasitology 132 , 49–56 (2006). Hurst, G. D. D., Jiggins, F. M. & Robinson, S. J. W. What causes inefficient transmission of male-killing Wolbachia in Drosophila? Heredity (Edinb) . 87 , 220–226 (2001). Kelleher, L. A. & Ramalho, M. O. Impact of rising temperatures on the bacterial communities of Aphaenogaster ants. Biol Open 14 , (2025). Bordenstein, S. R. & Bordenstein, S. R. Temperature Affects the Tripartite Interactions between Bacteriophage WO, Wolbachia, and Cytoplasmic Incompatibility. PLoS One . 6 , e29106 (2011). Charlesworth, J., Weinert, L. A., Araujo, E. V., Welch, J. J. & Wolbachia Cardinium and climate: an analysis of global data. Biol. Lett. 15 , 20190273 (2019). Frost, C. L., Mitchell, R., Smith, J. E. & Hughes, W. O. H. Genotypes and phenotypes in a Wolbachia -ant symbiosis. PeerJ 12, e17781 (2024). Noh, P. et al. Association between host wing morphology polymorphism and Wolbachia infection in Vollenhovia emeryi (Hymenoptera: Myrmicinae). Ecol. Evol. 10 , 8827–8837 (2020). Kelleher, L. A. & Ramalho, M. O. Impact of Species and Developmental Stage on the Bacterial Communities of Aphaenogaster Ants. Curr. Microbiol. 82 , 157 (2025). Ahmed, M. Z., Breinholt, J. W. & Kawahara, A. Y. Evidence for common horizontal transmission of Wolbachia among butterflies and moths. BMC Evol. Biol. 16 , 118 (2016). Ramalho, M., de Kim, O., Wang, Z., Moreau, C. S. & S. & Wolbachia Across Social Insects: Patterns and Implications. Ann. Entomol. Soc. Am. 114 , 206–218 (2021). Tolley, S. J. A., Nonacs, P. & Sapountzis, P. Wolbachia Horizontal Transmission Events in Ants: What Do We Know and What Can We Learn? Front Microbiol 10 , (2019). Hague, M. T. J. et al. Temperature effects on cellular host-microbe interactions explain continent-wide endosymbiont prevalence. Curr. Biol. 32 , 878–888e8 (2022). Hurst, G. D. D., Johnson, A. P., Schulenburg, J. H. & Fuyama, Y. G. v d Male-Killing Wolbachia in Drosophila: A Temperature-Sensitive Trait With a Threshold Bacterial Density. Genetics 156, 699–709 (2000). MOUTON, L., HENRI, H., BOULETREAU, M. & VAVRE, F. Effect of temperature on Wolbachia density and impact on cytoplasmic incompatibility. Parasitology 132 , 49–56 (2006). Wenseleers, T., Sundström, L. & Billen, J. Deleterious Wolbachia in the ant Formica truncorum. Proc. R Soc. Lond. B Biol. Sci. 269 , 623–629 (2002). Cariou, M., Duret, L. & Charlat, S. The global impact of Wolbachia on mitochondrial diversity and evolution. J. Evol. Biol. 30 , 2204–2210 (2017). Nelson, B. Y., INCREASED FUNGAL DIVERSITY ASSOCIATED & WITH APHAENOGASTER SPP. MORE EVIDENCE FOR KEYSTONE MUTUALISMS (Western Carolina University, 2012). University, M. N. S. W. C. & undefined Effects of myrmecochore species abundance, diversity, and fruiting phenology on Aphaenogaster (Hymenoptera: Formicidae) nesting and foraging in Southern. core.ac.ukMN SchultzWestern Carolina University, 2014•core.ac.uk (2014). (2014). Ness, J. H., Morin, D. F. & Giladi, I. Uncommon specialization in a mutualism between a temperate herbaceous plant guild and an ant: Are Aphaenogaster ants keystone mutualists? Oikos 118, 1793–1804 (2009). Umphrey, G. J. Morphometric discrimination among sibling species in the fulva – rudis – texana complex of the ant genus Aphaenogaster (Hymenoptera: Formicidae). Can. J. Zool. 74 , 528–559 (1996). Paysen, E. S. et al. FORMICIDAE) NESTING AND FORAGING IN SOUTHERN APPALACHIAN RICH COVE FORESTS. Ecology 10 , 5–7 (2018). Lenoir, A. & Devers, S. Alkaloid secretion inhibited by antibiotics in Aphaenogaster ants. C R Biol. 341 , 358–361 (2018). Pagalilauan, A., Pavloudi, C., Ospina, S. M., Smith, A. & Saw, J. Interaction with refuse piles is associated with co-occurrence of core gut microbiota in workers of the ant Aphaenogaster picea. Preprint at. https://doi.org/10.1099/acmi.0.000832.v3 (2025). Perfecto, I., Philpott, S. M. & Ants (Hymenoptera: Formicidae) and ecosystem functions and services in urban areas: a reflection on a diverse literature. Myrmecol News (2023). Tables Table 1. Results of paired sample Wilcox tests comparing average relative abundance of Wolbachia from ants where red, purple and/or blue clade ants co-occur. Comparison V-statistic p-value blue:red 10 0.0625 . blue:purple 9 0.125 purple:red 2 0.5 red+purple:blue 28 0.007813** yellow:green 37 0.049* red+purple/yellow:red+purple/green 9 0.125 Additional Declarations No competing interests reported. Supplementary Files Malagonetal092525SI.docx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 02 Mar, 2026 Reviews received at journal 17 Feb, 2026 Reviews received at journal 15 Feb, 2026 Reviewers agreed at journal 03 Feb, 2026 Reviewers agreed at journal 02 Feb, 2026 Reviewers agreed at journal 01 Feb, 2026 Reviewers agreed at journal 29 Jan, 2026 Reviewers agreed at journal 29 Jan, 2026 Reviewers invited by journal 27 Jan, 2026 Editor invited by journal 07 Jan, 2026 Editor assigned by journal 05 Jan, 2026 Submission checks completed at journal 05 Jan, 2026 First submitted to journal 04 Jan, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8515672","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":581795102,"identity":"853260cc-f3c8-4d88-8aa8-b5498ffee988","order_by":0,"name":"Daniel Malagon","email":"","orcid":"","institution":"Clemson University","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Malagon","suffix":""},{"id":581795107,"identity":"2be5f75d-07e3-4d96-927a-c5b2dabb5daa","order_by":1,"name":"Benjamin Camper","email":"","orcid":"","institution":"Clemson University","correspondingAuthor":false,"prefix":"","firstName":"Benjamin","middleName":"","lastName":"Camper","suffix":""},{"id":581795109,"identity":"e8fe45bc-8b09-4a7d-8596-432d21db4be8","order_by":2,"name":"Sophie Millard","email":"","orcid":"","institution":"Clemson University","correspondingAuthor":false,"prefix":"","firstName":"Sophie","middleName":"","lastName":"Millard","suffix":""},{"id":581795110,"identity":"00972887-9ec9-4d24-83d5-3579da02aaa5","order_by":3,"name":"Ernesto Recuero","email":"","orcid":"","institution":"Clemson University","correspondingAuthor":false,"prefix":"","firstName":"Ernesto","middleName":"","lastName":"Recuero","suffix":""},{"id":581795111,"identity":"05e98160-b897-4881-ac87-6660970c7841","order_by":4,"name":"Michael Caterino","email":"","orcid":"","institution":"Clemson University","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"","lastName":"Caterino","suffix":""},{"id":581795112,"identity":"4e208d2a-bb8a-4816-8501-e669d1eb2d17","order_by":5,"name":"Maslyn Greene","email":"","orcid":"","institution":"Clemson University","correspondingAuthor":false,"prefix":"","firstName":"Maslyn","middleName":"","lastName":"Greene","suffix":""},{"id":581795113,"identity":"0fcaa97b-1a84-47b5-b9c7-a66dd0e71171","order_by":6,"name":"Anna Seekatz","email":"","orcid":"","institution":"Clemson University","correspondingAuthor":false,"prefix":"","firstName":"Anna","middleName":"","lastName":"Seekatz","suffix":""},{"id":581795114,"identity":"802a3773-7090-425b-9fee-b61c0270ca07","order_by":7,"name":"Seth Bordenstein","email":"","orcid":"","institution":"Pennsylvania State University","correspondingAuthor":false,"prefix":"","firstName":"Seth","middleName":"","lastName":"Bordenstein","suffix":""},{"id":581795115,"identity":"05f99522-0614-42a9-85ef-ea54fc06ba00","order_by":8,"name":"Sarah Bordenstein","email":"","orcid":"","institution":"Pennsylvania State University","correspondingAuthor":false,"prefix":"","firstName":"Sarah","middleName":"","lastName":"Bordenstein","suffix":""},{"id":581795116,"identity":"3a989e77-04aa-4b41-a8d2-ed19bc2bba69","order_by":9,"name":"Sharon Bewick","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA00lEQVRIiWNgGAWjYDACCSBOYEiQY2CHCyUQp8WYgZkkLUBViQ1Ea5Gf3XxM4kFNWnp/M/PRjT8q7jDws+cY4NVicOdYmkTCsZzcGYfZ0m7znHnGINnzhoAWiRyzG4kNFbkbmHnMbjO2HWYwuEHAFvkZ+d9AWtINmPm/3fwJ1GJPSAvDjRw2oJacBANmHrYbvCBbJAj7xfxHwrE0Q6BfzIB+OcwjceZZAX6HzW5+bPijJlmev7352c0fFYfl+NuTN+B3GDrgIU35KBgFo2AUjAKsAADapEkh9JCS9AAAAABJRU5ErkJggg==","orcid":"","institution":"Clemson University","correspondingAuthor":true,"prefix":"","firstName":"Sharon","middleName":"","lastName":"Bewick","suffix":""}],"badges":[],"createdAt":"2026-01-04 23:23:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8515672/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8515672/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":101402942,"identity":"da919497-14c9-431a-8576-c8442d6093d4","added_by":"auto","created_at":"2026-01-29 10:12:50","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":151435,"visible":true,"origin":"","legend":"\u003cp\u003e(A) \u003cem\u003eWolbachia\u003c/em\u003erelative abundance (black: \u0026lt;0.5%, dark red: \u0026gt;0.5% and \u0026lt; 5%, red: \u0026gt;5%) for all 68 \u003cem\u003eAphaenogaster\u003c/em\u003e microbiomes shown based on site of collection (note that site coordinates are jittered to make all samples visible); (B) Moran’s I scatterplot for \u003cem\u003eWolbachia\u003c/em\u003e relative abundance; (C) Beta regression model of \u003cem\u003eWolbachia\u003c/em\u003e relative abundance as a function of the PC1 environmental axis.\u003c/p\u003e","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8515672/v1/af5f7a533ed736258c65dd2f.png"},{"id":101402954,"identity":"d5a5bae5-6bb5-498c-8612-5ac8e592e46f","added_by":"auto","created_at":"2026-01-29 10:12:52","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":146371,"visible":true,"origin":"","legend":"\u003cp\u003e(A) Violin plot of \u003cem\u003eWolbachia\u003c/em\u003e relative abundance for ants with the red clade COI sequence and the green clade CAD sequence (R-G, red/green), the red clade COI sequence and the yellow clade CAD sequence (R-Y, red/yellow), the purple clade COI sequence and the green clade CAD sequence (P-G, purple/green), the purple clade COI sequence and the yellow clade CAD sequence (P-Y, purple/yellow), the blue clade COI sequence and the green clade CAD sequence (B-G, blue/green) and the blue clade COI sequence and the yellow clade CAD sequence (B-Y, blue/yellow); (B) cophylogeny of the \u003cem\u003eAphaenogaster \u003c/em\u003eCOI sequences (left) and CAD sequences (right); COI (red, purple, blue) and CAD (green, yellow, black) clades are shown based on the color of the leaf nodes, while \u003cem\u003eWolbachia\u003c/em\u003e relative abundance (blue: low, red: high) is shown based on the color of the links; (C) results of a beta regression for the model Wolbachia ~ COI_clade*CAD_clade.\u003c/p\u003e","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8515672/v1/1eec18e30fa3240c9427e404.png"},{"id":101402924,"identity":"1f616703-d3ad-47f6-b4b4-9565ad1d3323","added_by":"auto","created_at":"2026-01-29 10:12:36","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":189181,"visible":true,"origin":"","legend":"\u003cp\u003eHeatmaps showing the number of reads of each \u003cem\u003eWolbachia\u003c/em\u003e ASV on each ant, organized according to host COI (A) and CAD (B) phylogenies and the \u003cem\u003eWolbachia\u003c/em\u003ephylogeny. The color scheme is logarithmic. COI and CAD clades are colored on their respective heat maps.\u003c/p\u003e","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8515672/v1/0b0fe67a387c0099a3d097ad.png"},{"id":101402938,"identity":"67d0b47b-cf5f-4d0f-9204-8e9c74d349de","added_by":"auto","created_at":"2026-01-29 10:12:49","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":68129,"visible":true,"origin":"","legend":"\u003cp\u003eMantel correlograms for sequence similarity based on the COI (A, C) and CAD (B, D) consensus trees, regressed against spatial distance (A, B) and environmental distance (C, D) and assessed with Spearman’s rank correlation. \u003cem\u003ep-\u003c/em\u003evalues shown on each panel reflect the significance of the corresponding Mantel tests (see also SI II Figures S2.2 and S2.3).\u003c/p\u003e","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8515672/v1/683bd436128c085b047a7a3a.png"},{"id":101402956,"identity":"1b921604-bc2e-40b2-a9bd-4c651885a9c2","added_by":"auto","created_at":"2026-01-29 10:12:56","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":370663,"visible":true,"origin":"","legend":"\u003cp\u003eCOI (A) and CAD (C) phylogenetic trees mapped to 14 sites of origin across 80 ants (A) and 92 ants (B) in GSMNP using the \u003cem\u003eplot.phylo.to.map\u003c/em\u003efunction in the phytools package. The node unassociated with a GSMNP location is the outgroup, \u003cem\u003eAphaenogaster umphreyi\u003c/em\u003e. Lines are colored based on whether the ant is included in the (A) red, blue or purple clades and (C) green, yellow or basal (white) clades. Pie charts show the fraction of ants at each location that come from each COI or CAD clade. (B, D) Violin plot showing comparison of values along the conglomerate environmental axis for red clade, blue clade and purple clade ants (B) and for the green and yellow clade ants (D).\u003c/p\u003e","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8515672/v1/6ab62877acdb4b2627245c11.png"},{"id":101402927,"identity":"af55500b-0569-4234-a40c-8cc75e5f165b","added_by":"auto","created_at":"2026-01-29 10:12:40","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":190878,"visible":true,"origin":"","legend":"\u003cp\u003eBeta regression models of total (A,B) \u003cem\u003eWolbachia\u003c/em\u003e relative abundance and relative abundances of the wArudA1 (C,F), wArudA2 (D,G) and wArudF1 (E,H) ASVs as a function of our conglomerate environmental variable (see Methods) for (A,C-E) red clade ants and (B,F-H) blue clade ants. p-values indicate significance of the environment, and specific sites are denoted as follows: Abram’s Creek (closed red circles), Tremont (closed upward red triangles), Cades Cove (closed squares), Twin Creeks (closed downward triangles), Oconaluftee (closed diamond), Trillium Gap (star), Snake Den (open upward triangle), Albright Grove (open downward triangle), Cataloochee (closed upward blue triangles), Purchase Knob (open squares), Brushy Mountain Myrtle (closed blue circles), Brushy Mountain (open circles).\u003c/p\u003e","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8515672/v1/bd3980a441471d8955a6bff1.png"},{"id":101402980,"identity":"a26cc283-9872-42da-862f-d02831902d31","added_by":"auto","created_at":"2026-01-29 10:13:02","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1558815,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8515672/v1/7b9698dc-a708-4121-89ce-78c1b09d25fd.pdf"},{"id":101402921,"identity":"ec8498bb-afcb-4886-b426-564a9fb5f8cd","added_by":"auto","created_at":"2026-01-29 10:12:35","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":10730157,"visible":true,"origin":"","legend":"","description":"","filename":"Malagonetal092525SI.docx","url":"https://assets-eu.researchsquare.com/files/rs-8515672/v1/a70d50d53680db25accd805e.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Host and Microbe Scale Processes Jointly Shape Spatial Variation in Aphaenogaster (Hymenoptera: Formicidae)associated Wolbachia","fulltext":[{"header":"Introduction","content":"\u003cp\u003eTwo of the most important drivers of a taxon\u0026rsquo;s spatial distribution are environmental selection and dispersal limitation.\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e Environmental selection (i.e., taxon sorting) causes spatial structure by limiting access of certain taxa to certain locations based on their ability to tolerate local biotic or abiotic conditions. This can yield a wide array of different spatial patterns, including the evolution of ecotypes, genetic isolation by environment (IBE),\u003csup\u003e2,3\u003c/sup\u003e and distance-decay of community similarity.\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e Like environmental selection, dispersal limitation also causes spatial structure by limiting access of certain taxa to certain locations.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e In the case of dispersal limitation, however, access is limited by the ability of taxa to cross physical barriers and/or unsuitable habitat.\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e Dispersal limitation can result in a wide variety of different spatial patterns, including genetic isolation by distance (IBD), metapopulation dynamics and, once again, distance-decay of community similarity.\u003c/p\u003e \u003cp\u003eWhile the spatial distributions of free-living organisms are primarily governed by environmental selection and dispersal limitation at the scale of the organism, the spatial distributions of host-associated (HA) microbes are governed by environmental selection and dispersal limitation at two scales: the scale of the microbe and the scale of their host. Thus, spatial variation in HA microbes is affected not only by dispersal of microbes from one host to another (microbe scale dispersal limitation) and differences in conditions on the host (microbe scale environmental selection; e.g., due to differences in host genetics or phenotype), but also by dispersal of hosts from one location to another (host scale dispersal limitation) and differences in conditions where the hosts reside (host scale environmental selection).\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e,\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e This makes interpretation of HA microbial spatial patterns challenging. More specifically, because host-scale (i.e., variation in landscape-level environment and host movement) and microbe-scale (i.e., variation in host characteristics, transfer between hosts) processes are often correlated, it can be difficult to disentangle their independent effects on HA microbes. As an example, host genetic distance often correlates with host geographic distance and/or host environmental distance due to processes like genetic drift and local adaptation of the host.\u003csup\u003e\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e Consequently, the environment external to the host (host-scale) and the environment created by the host (microbe-scale) frequently co-vary, making it difficult to attribute spatial patterns to host- versus microbe-scale environmental selection.\u003c/p\u003e \u003cp\u003eFurther complicating the interpretation of spatial patterns in HA microbes is the potential for bi-directional interactions across scales. Thus, just as host scale processes can impact the spatial ecology of HA microbes (i.e., top-down effects), microbe scale processes can influence the spatial ecology of their hosts (i.e., bottom-up effects).\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e Many microbes, for instance, alter host dispersal, either as a side-effect of their impacts on host health\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e or because they have evolved to manipulate host dispersal to promote their own transmission.\u003csup\u003e1314\u003c/sup\u003e HA microbes can also alter host environmental selection. In some systems, for example, it is the abiotic tolerances of obligate endosymbionts, rather than the hosts themselves that restrict environmental associations.\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e Beyond dispersal and selection, HA microbes can also manipulate host reproduction. This is important to spatial patterns because it can impact host gene flow between regions.\u003csup\u003e161718\u003c/sup\u003e One of the most notorious examples is \u003cem\u003eWolbachia\u003c/em\u003e\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e \u0026ndash; a widespread\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e bacterium that can affect host genetics by inducing cytoplasmic incompatibility (CI) among hosts harboring differing \u003cem\u003eWolbachia\u003c/em\u003e strains.\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,2223,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eBecause of the complicated, correlated and often bi-directional nature of interactions between host- and microbe-scale processes, it can be difficult to determine the degrees to which processes at each scale influence the spatial structure of HA microbes. Regions where cross-scale correlations break down, however, can sometimes be useful to disentangle the effects of host- versus microbe-scale processes. In this paper, we investigate the roles of host- and microbe-scale spatial processes on spatial patterns in \u003cem\u003eWolbachia\u003c/em\u003e carriage among \u003cem\u003eAphaenogaster fulva-rudis-texana\u003c/em\u003e (Hymenoptera: Formicidae; henceforth \u0026lsquo;\u003cem\u003eA. rudis\u003c/em\u003e\u0026rsquo;) complex ants in Great Smoky Mountains National Park (GSMNP). Leveraging previously derived, high resolution GSMNP bioclimatic variables,\u003csup\u003e25,26\u003c/sup\u003e we assess the extent to which \u003cem\u003eWolbachia\u003c/em\u003e relative abundance varies as a function of the environment of the host (i.e., host scale environmental selection) and/or spatial distance (i.e., host scale dispersal limitation). We then assess the extent to which \u003cem\u003eWolbachia\u003c/em\u003e relative abundance varies as a function of host genetics (i.e., microbe scale environmental selection) and/or host mitotype (i.e., microbe scale dispersal limitation). Next, we show how host- and microbe-scale processes are strongly correlated, potentially complicating interpretation of spatial patterns. Finally, we repeat our analyses, subsetting our dataset in ways that allow us to break correlations between host- and microbe-scale processes. Ultimately, this allows us to identify the independent roles of processes at both scales as determinants of \u003cem\u003eWolbachia\u003c/em\u003e relative abundance in \u003cem\u003eAphaenogaster\u003c/em\u003e ants from GSMNP.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy Species\u003c/h2\u003e \u003cp\u003eAnts of the \u003cem\u003eA. rudis\u003c/em\u003e complex are abundant and widespread in eastern hardwood forests of North America.\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e With a generalist strategy that includes colonization of a wide array of forest types,\u003csup\u003e28\u003c/sup\u003e and broad latitudinal,\u003csup\u003e29\u003c/sup\u003e and elevational ranges,\u003csup\u003e30\u0026ndash;34\u003c/sup\u003e this complex is ideal for examining effects of host scale environmental selection on HA microbes. Further, \u003cem\u003eA. rudis\u003c/em\u003e complex ants within GSMNP comprise four incipient species - \u003cem\u003eA. rudis\u003c/em\u003e, \u003cem\u003eA. picea\u003c/em\u003e, \u003cem\u003eA. fulva\u003c/em\u003e, and \u003cem\u003eA. carolinensis\u003c/em\u003e. Generally speaking, \u003cem\u003eA. picea\u003c/em\u003e is found at higher elevations while \u003cem\u003eA. rudis\u003c/em\u003e is found at lower elevations.\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e Despite these differences,\u003csup\u003e35,36\u003c/sup\u003e members of the complex co-occur extensively, both within GSMNP and across the Southeast broadly, \u003csup\u003e37\u003c/sup\u003e possibly because they are in the early stages of lineage divergence.\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e This makes the \u003cem\u003eA. rudis\u003c/em\u003e complex ideal for examining the effects of microbe scale environmental selection (i.e., host genetics) on HA microbes.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eStudy Sites\u003c/h3\u003e\n\u003cp\u003eAll ants were collected from GSMNP. GSMNP is an ideal location for studying the effects of host scale environmental selection due to its mountainous terrain. In particular, the large elevational gradients characteristic of the GSMNP landscape result in significant bioclimatic variation over relatively small distances.\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e,\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e \u003cem\u003eA. rudis\u003c/em\u003e complex ants were collected from 12 All Taxa Biodiversity Index (ATBI) plots\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e as well as an additional plot at one of the lowest elevations in the park (13 total sites) from Fall 2019 - Summer 2021. These plots vary greatly in elevation (522-1,494 meters) and were selected to represent the dominant ecosystems (i.e., distinct environments with different selection pressures) present in GSMNP. At each plot, we collected between 4 and 15 1m x 1m quadrats of litter across 6 to 11 unique time-points, depending on site accessibility. All quadrats within any sampling period were selected to be at least 5m apart to avoid sampling the same colony within multiple quadrats, and at most one ant from any quadrat was selected for sequencing. In addition to quadrat sampling, we spent\u0026thinsp;~\u0026thinsp;30 minutes searching for and aspirating ants during each visit. (For additional details regarding ant collection and identification, see SI I).\u003c/p\u003e\u003cp\u003e\u003cem\u003eCOI and CAD Gene Extraction, Amplification, and Sequencing\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eGenomic DNA was extracted from 1 leg of each ant using the Thermo Scientific Genejet Genomic DNA purification kit following the manufacturer\u0026rsquo;s protocol. We amplified the COI gene following Harden et al. 2022\u003csup\u003e39\u003c/sup\u003e using the primers LCO1490 and HCO2198 in 25-\u0026micro;L PCR reactions containing 2.5 \u0026micro;L of template DNA, 1.0 \u0026micro;L of each primer, 2.5 \u0026micro;L of dNTPs, 2.0 \u0026micro;L of MgCl2, 0.125 \u0026micro;L of Platinum Taq, 2.5 \u0026micro;L of buffer, and 13.375 \u0026micro;L of Invitrogen ultra-pure water (Catalog number: 10977023). Reactions were carried out using an Eppendorf MasterCycler with the following settings: initial denaturation stage of 180 seconds at 95 \u0026deg;C followed by 35 cycles of a denaturation stage at 94 \u0026deg;C for 30 seconds, an annealing stage at 45 \u0026deg;C for 30 seconds, an extension stage at 72 \u0026deg;C for 45 seconds, and ending with a final extension at 72 \u0026deg;C for 180 seconds. CAD was amplified\u0026nbsp;\u0026nbsp;using primers CD439F and CD851R.\u003csup\u003e51\u0026nbsp;\u003c/sup\u003eCAD\u0026nbsp;reactions were carried out as above using an Eppendorf MasterCycler beginning with an initial denaturation at 95\u0026deg;C for 3 minutes followed by 40 cycles of a denaturation stage at 94\u0026deg;C for 30 seconds, an annealing stage at 50 \u0026deg;C for 30 seconds, an extension stage at 72 \u0026deg;C for 50 seconds, and a final extension at 72 \u0026deg;C for 5 minutes. Amplified genes were forward and reverse Sanger sequenced at Psomagen.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCOI and CAD Phylogenetic Tree Construction\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eForward and reverse COI and CAD Sanger sequencing reads were merged in R (v.4.2.1)\u003csup\u003e40\u003c/sup\u003e using the package sangeranalyseR\u0026nbsp;(v.1.6.1).\u003csup\u003e41\u003c/sup\u003e After quality control, 91 CAD sequences and 80 COI sequences remained from our original 102 samples. We aligned merged reads, without a reference, using our own data as well as data from an \u003cem\u003eA. umphreyi\u003c/em\u003e individual that we chose as our outgroup (GenBank Accessions: KP860492.1 and KP730141.1).\u003csup\u003e42\u003c/sup\u003e COI and CAD genes were aligned separately using the MAFFT algorithm in Mesquite. We then employed maximum likelihood methods for phylogenetic tree construction through the web server version (v. 2.4.0) of IQ-Tree\u003csup\u003e43\u003c/sup\u003e using default parameters. This was repeated for the COI and CAD single alignments separately to produce single gene trees. For the COI (mtDNA marker) and CAD (nuDNA marker) genes, the best-fit models of nucleotide substitution, selected based on BIC and using ModelFinder within IQ-Tree, were TPM2u+F+R2 and K2P+R2 respectively. Tree discordance between COI and CAD gene trees was calculated using the \u003cem\u003ecospeciation\u003c/em\u003e function and visualized using the \u003cem\u003ecophylo\u003c/em\u003e function both in the \u003cem\u003ephytools\u003c/em\u003e package in R (version 4.2.1). All statistics were performed using separate COI and CAD trees rather than a concatenated tree. For host-specific analyses, we used all COI and/or CAD sequences available (80 and 91 samples respectively). For analyses that paired COI/CAD and \u003cem\u003eWolbachia\u003c/em\u003e, we only used samples for which both host and \u003cem\u003eWolbachia\u003c/em\u003e data were available (46 samples for COI/\u003cem\u003eWolbachia\u003c/em\u003e, 52 samples CAD/\u003cem\u003eWolbachia\u003c/em\u003e, 41 samples COI/CAD/\u003cem\u003eWolbachia\u003c/em\u003e). For all analyses, we used a square-root transformation of branch lengths\u003csup\u003e44\u003c/sup\u003e in either the COI or CAD trees.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eExtraction, Amplification, and Sequencing of Microbial DNA\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eMicrobial DNA was extracted using the ZymoBIOMICS DNA Microprep kit (catalogue number D3401). Prior to extraction, individual ants were surface sterilized for 1 minute in 10% bleach solution and rinsed 3 times in ultra-pure water to ensure transient microbes on the exoskeleton were removed. We performed the extraction following the manufacturer\u0026rsquo;s protocol with the following modifications: 600 \u0026micro;L input of ant lysate was added to 1800 \u0026micro;L of binding buffer due to low input. We skipped filtering steps 4, 12, and 13 due to low input and heated the elution buffer to 60 \u0026deg;C prior to elution to maximize DNA concentrations. Samples were lysed for 40 minutes using a vortex genie and for 20 minutes using a tissuelyzer2 (Qiagen) at 25 x/second to effectively lyse the ant exoskeleton.\u003c/p\u003e\n\u003cp\u003eWe characterized entire bacterial communities by amplifying the V4 region of the 16S rRNA gene using touchdown PCR with common dual index primers.\u003csup\u003e45\u003c/sup\u003e Briefly, we amplified in 20-\u0026micro;L PCR reactions containing 5 \u0026micro;L DNA, 5 \u0026micro;L of the primer set (515F and 806R), 2 \u0026micro;L of 10x Accuprime PCR Buffer II, 0.15 \u0026micro;L of Accuprime HiFi polymerase, and 7.85 \u0026micro;L of ultra-pure water per reaction. Reactions were carried out using an Eppendorf MasterCycler with the following settings: an initial denaturation stage of 120 seconds at 95 \u0026deg;C followed by 20 cycles of a denaturation stage at 95 \u0026deg;C for 20 seconds, an annealing stage at 60 \u0026deg;C for 20 seconds, and an extension stage at 72 \u0026deg;C for 5 minutes. This was followed by 20 cycles of a denaturation stage at 95 \u0026deg;C for 20 seconds, an annealing stage at 55 \u0026deg;C for 20 seconds, an extension stage at 72 \u0026deg;C for 5 minutes, and a final extension stage at 72 \u0026deg;C for 10 minutes. Samples were normalized prior to pooling using a SequalPrep Normalization Plate Kit (catalogue number #A1051001). Pooled sample library concentrations were quantified via KAPA qPCR (Roche, cat. #07960140001). Multiplexed pooled libraries were sequenced paired end 2 x 300 cycles on an Illumina NextSeq 2000 to an average depth of 245,448 reads. Resulting FASTQ files were analyzed using a custom Qiime2\u003csup\u003e46\u003c/sup\u003e pipeline. (For additional information on Taxonomic Assignment and Dataset Manipulation, see SI I; see also Github repository below).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eEnvironmental axis\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe considered elevation, mean soil moisture (volumetric content), soil minimum temperature, and soil maximum temperature derived from previous GSMNP studies.\u003csup\u003e25,26\u003c/sup\u003e We first z-transformed these four environmental variables, and then calculated environmental distances based on the Euclidean distances between the z-transformed environmental variable values at each capture location. Next, we performed a principal component analysis (PCA) based on the environmental variables associated with each ant. We then selected the first principal component (PC1; i.e., the axis capturing the most variation in the system) and used the value of each ant along this axis for subsequent analysis of environmental effects.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eStatistical Analyses\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAll statistical analyses were performed in R (v.4.2.1).All data and code can be found in the following github repository: https://github.com/dmalago/Aphaenogaster_GSMNP_Microbiome.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eSpatial autocorrelation\u003c/u\u003e: \u0026nbsp; We used the \u003cem\u003ejitter\u0026nbsp;\u003c/em\u003efunction from base R tojitter site values for latitude and longitude so that each ant sample had a unique location. We then used the \u003cem\u003echooseCN\u003c/em\u003e function from the \u003cem\u003eadespatial\u003c/em\u003e package to create a connection network. This was done using Delaunay triangulation (type 1). From this network, we used the \u003cem\u003enbdists\u0026nbsp;\u003c/em\u003efunction from the \u003cem\u003espdep\u003c/em\u003e package to calculate edge distances. Next, we used the \u003cem\u003enb2listw\u003c/em\u003e function from the \u003cem\u003espdep\u003c/em\u003e package to find row-standardized (style = 4) spatial weights for network edges. Finally, we inputted \u003cem\u003eWolbachia\u003c/em\u003e relative abundances and the list of spatial weights into the \u003cem\u003emoran.test\u003c/em\u003e function, also from the \u003cem\u003espdep\u003c/em\u003e package, to determine whether there was significant autocorrelation. Results were visualized using the \u003cem\u003emoran.plot\u003c/em\u003e function, again from the \u003cem\u003espdep\u003c/em\u003e package.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eRegression Analyses\u003c/u\u003e: We used generalized linear models\u003csup\u003e47\u003c/sup\u003e with beta distributions to regress \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003erelative abundances against principal component one (PC1) of the environmental PCA (\u003cem\u003eWolbachia ~ Environment\u003c/em\u003e, see above for axis definition). This was done for all ants and for ants from individual lineages separately. We also regressed \u003cem\u003eWolbachia\u003c/em\u003e relative abundance against host COI and CAD clade/group (\u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003e~ COI\u003csub\u003eClade\u003c/sub\u003e*CAD\u003csub\u003eClade\u003c/sub\u003e, see \u003cem\u003eResults\u003c/em\u003e for information on clades/groups). All regressions were performed using the\u003cem\u003e\u0026nbsp;betareg\u003c/em\u003e function from the \u003cem\u003ebetareg\u003c/em\u003e package.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eHeatmaps\u003c/u\u003e: We constructed heatmaps of \u003cem\u003eWolbachia\u003c/em\u003e relative abundance for each \u003cem\u003eWolbachia\u003c/em\u003e ASV in each ant using the \u003cem\u003epheatmap\u003c/em\u003e function from the \u003cem\u003epheatmap\u003c/em\u003e package. This was done ordering ants according to both the COI and the CAD phylogenies.\u003c/p\u003e\n\u003cp\u003e\u003cu\u003eMantel tests\u003c/u\u003e: We used Mantel\u003csup\u003e48\u003c/sup\u003e tests to assess whether host relatedness varied with geographic distance and whether host relatedness varied with environmental similarity. All Mantel tests were performed using the \u003cem\u003emantel\u003c/em\u003e function from the \u003cem\u003evegan\u003c/em\u003e package(v. 2.6.4).\u003csup\u003e49\u003c/sup\u003e We then used the \u003cem\u003emantel.correlog\u003c/em\u003e function, also from the \u003cem\u003evegan\u003c/em\u003e package (v. 2.6.4), to calculate mantel correlations at varying geographic or environmental distance classes. For spatial distance, we used the Haversine formula to calculate the great circle distance between all ant capture locations based on jittered (see above) longitude and latitude. This was implemented using the \u003cem\u003edistm\u0026nbsp;\u003c/em\u003efunction from the \u003cem\u003egeosphere\u0026nbsp;\u003c/em\u003epackage (v. 1.5-20). For environmental distance, we considered the distance between ants along our PC1 environmental axis (see \u003cem\u003eEnvironmental axis\u003c/em\u003e, above). For host relatedness, we used distances based on the branch lengths of the square root transformed COI or CAD trees. Branch lengths were calculated using the \u003cem\u003ecophenetic.phylo\u0026nbsp;\u003c/em\u003efunction from the \u003cem\u003eape\u003c/em\u003e package (v. 5.7.1).\u003csup\u003e50\u003c/sup\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cu\u003ePaired sample Wilcoxon tests:\u0026nbsp;\u003c/u\u003eWe used paired sample Wilcoxon tests\u003csup\u003e51\u003c/sup\u003e to test for differences in \u003cem\u003eWolbachia\u003c/em\u003e relative abundance between host clades/groups (see \u003cem\u003eResults\u003c/em\u003e), controlling for environment/spatial autocorrelation. First, we identified sites where different COI clades/groups were co-localized. We then averaged \u003cem\u003eWolbachia\u003c/em\u003e relative abundances for all ants from each clade/group at each site of co-localization. This gave a single \u003cem\u003eWolbachia\u003c/em\u003e relative abundance for each clade/group at each site. We then used these average \u003cem\u003eWolbachia\u003c/em\u003e relative abundances in paired sample Wilcoxon tests, pairing ants from different clades/groups at the same sites.\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cem\u003eAnt Clades/Groups\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eUsing the COI consensus tree (Fig. S2.1 – left side), we visually identified 2 main clades (henceforth the “blue” and “purple” clades) as well as a set of basal ants (henceforth the “red” [paraphyletic] group, see also SI Figure S3.6B for haplotype maps suggesting these are a separate group but not necessarily basal to the blue and purple clades). Using the CAD consensus tree (Fig. S2.1 – right side), we visually identified two main clades (henceforth the “green” and “yellow” clades) as well as several basal ants (henceforth the “black” group). Although consensus trees for COI and CAD sequences displayed significantly different topologies (RF distance = 148, p = 1), there was a relatively high degree of congruence between the red, purple and blue COI clades/groups and the green and yellow CAD clades. In particular, ants with red group COI sequences were significantly more likely to have green clade CAD sequences, while ants with blue clade COI sequences were significantly more likely to have yellow clade CAD sequences (Fisher’s exact test: \u003cem\u003ep =\u0026nbsp;\u003c/em\u003e0.00067, see SI II Figure S2.2, Table S2.1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAnt Microbiota\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eBased on regression analyses (see \u003cem\u003eStatistical Analyses\u003c/em\u003e), \u003cem\u003eWolbachia\u003c/em\u003e was the only microbial taxon whose relative abundance varied significantly across either environment or host clade (see SI III Figures S3.1 and S3.2). Consequently, we chose to focus on elucidating spatial patterns in \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ecarriage. Thirty-two separate ASVs were taxonomically classified to the \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003egenus, including members of supergroups A, B and F (see SI III Figure S3.3).\u003csup\u003e18\u003c/sup\u003e Only three, however, independently comprised more than 0.01% of reads across all ant microbiota (two from supergroup A, and one from supergroup F). These were also the only \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eASVs found in more than one ant.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eHost-scale variation in Wolbachia\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eWolbachia\u003c/em\u003e relative abundances exhibited significant spatial variation across GSMNP (see Figure 1). Whereas\u0026nbsp;\u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ewas present on most ants from the western side of the park,\u0026nbsp;\u003cem\u003eWolbachia\u003c/em\u003e was absent or present in very low abundances (\u0026lt;0.05%, see Figure 1A) on most ants from the eastern side of the park. This resulted in significant positive spatial autocorrelation in \u003cem\u003eWolbachia\u003c/em\u003e relative abundance (see Figure 1B). Regressing \u003cem\u003eWolbachia\u003c/em\u003e relative abundance against the PC1 environmental axis further indicated that \u003cem\u003eWolbachia\u003c/em\u003e relative abundance was higher in ants from warmer, drier, lower elevation regions of the park and that \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003erelative abundance was lower in ants from cooler, wetter, higher elevation regions of the park (see Figure 1C). This was consistent with overall spatial trends, since the western side of the park is generally warmer, drier and lower elevation than the eastern side of the park. Using variance partitioning to separate the spatial versus environmental drivers of \u003cem\u003eWolbachia\u003c/em\u003e abundance, we found that both spatial distance (9% of variation) and the interaction between spatial distance and environment (16% of variation) were important for explaining differences in \u003cem\u003eWolbachia\u003c/em\u003e relative abundances across ants (see SI III Figure S3.5). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eMicrobe-scale variation in Wolbachia\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eWolbachia\u003c/em\u003e relative abundance varied significantly with both mitochondrial and nuclear sequence variation of the host (see Figure 2). More specifically, our regression model suggested that ants with red group and purple clade COI sequences had significantly higher \u003cem\u003eWolbachia\u003c/em\u003e relative abundances as compared to ants with blue clade COI sequences, and that ants with red group and (marginally) purple clade COI sequences had significantly lower \u003cem\u003eWolbachia\u003c/em\u003e relative abundances when they had yellow as compared to green clade CAD sequences. Closer inspection suggested that three distinct \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eASVs were responsible for the majority of \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ein our system (see Figure 3, see also SI III figure S3.6). Two of the three dominant ASVs (henceforth wArudA1, wArudA2) were closely related, clustering together insupergroup A and differing from each other by 1/214 base pairs (bp). \u0026nbsp;By contrast, the third dominant ASV (henceforth wArudF1) was more distantly related, clustering in supergroup F and differing from wArudA1 and wArudA2 by 8/214 and 9/214 bp respectively (see SI III, Figure S3.3). Notably, these three ASVs showed strong variation across COI ant clades, indicating a likely link between mitotype and the presence of specific \u003cem\u003eWolbachia\u003c/em\u003e variants (see Figure 3A, see also SI III Figures S3.6B and S3.7). In particular, whereas ants from the red clade had all three \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eASVs, ants from the purple clade had only wArudA2, and ants from the blue clade had none of the three \u003cem\u003eWolbachia\u003c/em\u003e ASVs. As compared to host COI sequences, the presence/absence of individual \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eASVs was not as strongly correlated with host CAD sequences. Nevertheless, consistent with the lower abundance of \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ein yellow clade ants overall, a larger proportion of green clade ants had all three \u003cem\u003eWolbachia\u003c/em\u003e ASVs, while a larger proportion of yellow clade ants had no \u003cem\u003eWolbachia\u003c/em\u003e ASVs (see Figure 3B, see also SI III Figures S3.6C and S3.8).\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCorrelation between host- and microbe-scale drivers of variation\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eMantel tests based on Spearman, but not Pearson correlation coefficients demonstrated a significant relationship between COI sequence similarity and geographic distance (Spearman: r: 0.1005, p = 0.020; Pearson: r = 0.0703; p = 0.082; Figure 4A, see also SI II Figure S2.3A,B). Likewise, Mantel tests based on both Spearman and Pearson correlation coefficients demonstrated a significant relationship between COI sequence similarity and environmental similarity (Spearman: r = 0.0912, p = 0.014; Pearson: r = 0.0685; p = 0.042; Figure 4C, see also SI II Figure S2.3C,D). Distance-based redundancy analysis (dbRDA) and variance partitioning supported results from individual Mantel tests, suggesting that both distance and distance + environment explained the similarity of COI sequences (see SI II Figure S2.5A). Consistent with the observed correlations between COI sequence similarity and both spatial and environmental distance, ants with red group, purple clade and blue clade COI sequences exhibited different distributions across GSMNP. Red group ants dominated lower elevations on the western side of the park, while blue clade ants dominated higher elevations on the eastern side of the park. Purple clade ants were predominantly found towards the center of the park where they co-occurred with the other two clades/groups across a range of elevations (see Figure 5A). In accordance with their differing distributions, environmental conditions were also significantly different across the three COI clades/groups (see Figure 5B; Kruskal-Wallis chi-squared = 26.681, df = 2, p-value = 1.6E-6, see also SI II Figure S2.6A). Ants from the red group were associated with warmer, drier, and lower elevation conditions while ants from the blue clade were associated with colder, moister and higher elevation conditions. Ants from the purple clade were intermediate, though differences between the blue and the purple clades were not significant.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eUnlike COI sequences, no significant relationships were observed between CAD sequence similarity and either geographic distance (Spearman: r: -0.0902, p = 0.978; Pearson: r = -0.0634 ; p = 0.882; Figure 4B, see also SI II Figure S2.4A,B) or environmental distance (Spearman: r = -0.0319, p =0.814 ; Pearson: r = -0.0379, p = 0.832; Figure 4D, see also SI II Figure S2.4C,D). Likewise, CAD sequences did not exhibit obviously different distributions across GSMNP (see Figure 5C), nor were environmental conditions significantly different between the green and yellow clades (see Figure 5D, Kruskal-Wallis chi-squared = 2.0236, df =1, p-value = 0.1549; see also SI II Figure S2.6B).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eIsolating host-scale drivers of variation in Wolbachia\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;We examined the role of environment (host-scale process) while controlling for COI sequence (microbe-scale process) by considering the effects of the PC1 environmental variable on \u003cem\u003eWolbachia\u003c/em\u003e carriage by blue clade and red group ants separately. When we did this, we found that overall \u003cem\u003eWolbachia\u003c/em\u003e abundance, as well as abundance of each of the three dominant \u003cem\u003eWolbachia\u003c/em\u003e ASVs (wArudA1, wArudA2, and wArudF1) varied in response to environment (see Figure 6A,C-E) for red group ants. In particular, there was a precipitous decline in \u003cem\u003eWolbachia\u003c/em\u003e at a threshold environmental PC1 value of approximately -2 (roughly corresponding to a minimum soil temperature of -5\u003csup\u003e°\u003c/sup\u003eC, a soil moisture of 0.073 % volumetric moisture content, a soil temperature range of 32.5\u003csup\u003e°\u003c/sup\u003eC and an elevation of 622 m) with the two ants from the two coldest sites (Oconaluftee (OCO), Trillium Gap (TG)) exhibiting much lower (\u0026lt;0.5%) \u003cem\u003eWolbachia\u003c/em\u003e relative abundance as compared to ants from other sites. While the ant from Trillium Gap was yellow clade for its CAD sequence (i.e., potentially lower \u003cem\u003eWolbachia\u003c/em\u003e due to host genetics), the ant from Oconaluftee had a basal (‘black group’) CAD sequence. Unlike red group ants, the relationship between \u003cem\u003eWolbachia\u003c/em\u003e and environment was less apparent for blue clade ants, only reaching significance for overall \u003cem\u003eWolbachia\u003c/em\u003e relative abundance and the abundance of the wArudA2 ASV and, even then, only reaching significance when an outlier ant from Ramsey Cascades was removed. Notably, this was a generally unusual ant containing both wArudA2 and wArudF1 (most other ants either contained all three \u003cem\u003eWolbachia\u003c/em\u003e ASVs or else only wArudA2, see SI III Figure S3.6). \u0026nbsp;The weaker relationship between \u003cem\u003eWolbachia\u003c/em\u003e and environment in blue clade ants is not surprising given the dramatically lower \u003cem\u003eWolbachia\u003c/em\u003e relative abundances in blue clade ants in general. Overall, within-clade comparisons, particularly for red group ants, suggest that host-scale processes (environmental conditions and/or host dispersal limitation) independently impact the spatial distribution of \u003cem\u003eAphaenogaster\u003c/em\u003e-associated\u003cem\u003e\u0026nbsp;Wolbachia\u003c/em\u003e across GSMNP.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eIsolating microbe-scale drivers of variation in Wolbachia\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;We examined the effects of COI sequence (microbe-scale process) while controlling for environment (host-scale process) by contrasting average \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ecarriage of ants from different COI clades/groups paired according to collection site. When we did this, we found that red group and blue clade ants co-occurred at four sites, and that the difference in \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ecarriage was borderline in significance. Likewise, we found that red group and purple clade ants and purple clade and blue clade ants co-occurred at two and three sites respectively, and that the differences in \u003cem\u003eWolbachia\u003c/em\u003e carriage were not significant. Finally, we found that red group + purple clade and blue clade ants co-occurred at seven sites, and that the difference in \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ecarriage was highly significant (see Table 1). In addition, we examined the effects of CAD sequence (microbe-scale process) while controlling for environment (host-scale process) by contrasting \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ecarriage across ants from different CAD clades paired according to collection site. Again, when we did this, we found that green and yellow clade ants co-occurred at nine sites, and that the difference in \u003cem\u003eWolbachia\u003c/em\u003e carriage was significant. Finally, we examined the effects of COI/CAD sequence while controlling for environment. In this case, we found that red + purple/yellow clade ants and red + purple/green clade ants co-occurred at four sites, and the difference in \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ecarriage was not significant. Overall, these comparisons across paired samples suggest that microbe-scale processes (host genetics and/or host-to-host transmission), particularly processes correlated with COI sequence, independently impact the spatial distribution of \u003cem\u003eAphaenogaster\u003c/em\u003e-associated \u003cem\u003eWolbachia\u003c/em\u003e across GSMNP.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this paper, we examined host- and microbe-scale patterns in \u003cem\u003eWolbachia\u003c/em\u003e relative abundance among \u003cem\u003eA. rudis\u0026nbsp;\u003c/em\u003ecomplex ants from across Great Smoky Mountains National Park (GSMNP). Our goal was to determine whether spatial structure was primarily attributable to host-scale processes, microbe-scale processes or a combination of the two (i.e., multiscale processes).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eIndependent effects of host-scale processes\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe inferred independent effects of host-scale processes – more specifically, host-scale environmental selection – based on how environment impacted \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eload across individual ant COI clades/groups. In particular, we found that some combination of low temperature, high moisture and high elevation lowers \u003cem\u003eWolbachia\u003c/em\u003e abundance/persistence in our system (see below). This is not unprecedented.\u003csup\u003e52,53\u003c/sup\u003e In a recent study of captive \u003cem\u003eA. rudis\u003c/em\u003e colonies, for example, \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003erelative abundance decreased in response to artificial warming.\u003csup\u003e54\u003c/sup\u003e While this trend is opposite what we observed, it nonetheless indicates that temperature can affect \u003cem\u003eWolbachia\u003c/em\u003e abundance. Indeed, differences in directionality could merely reflect the existence of an optimum in the \u003cem\u003eWolbachia\u003c/em\u003e thermal performance curve. Notably, ants in the captive study were warmed to a constant temperature of 32°C, which is much higher than the ambient annual temperatures experienced by even the lowest elevation ants in our system (though not necessarily higher than the maximum annual temperature at the warmest sites). Thus, our range of temperatures may have been below the optimal growth temperature for \u003cem\u003eAphaenogaster\u003c/em\u003e-associated \u003cem\u003eWolbachia\u003c/em\u003e, causing an increase in \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ewith temperature. By contrast, the captive study may have heated ants past the optimal growth temperature, causing a decrease in \u003cem\u003eWolbachia\u003c/em\u003e with temperature. This is in keeping with studies of other Hymenoptera species\u003csup\u003e55\u003c/sup\u003e which have demonstrated reduced \u003cem\u003eWolbachia\u003c/em\u003e carriage (e.g., due to temperature dependent interactions with a \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ephage) at both excessively low and excessively high temperatures. Interestingly, a recent meta-analysis\u003csup\u003e56\u003c/sup\u003e across ~2500 species of arthropods also hints at the possibility of a unimodal response of \u003cem\u003eWolbachia\u003c/em\u003e to temperature, with \u003cem\u003eWolbachia\u003c/em\u003e prevalence increasing with temperature in temperate zones (i.e., from a lower baseline temperature) and decreasing with temperature in tropical zones (i.e., from a higher baseline temperature).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;While the host-scale patterns that we observed are well explained by host-scale environmental selection, they are not as easily explained by host-scale dispersal limitation. If host-scale dispersal limitation was fully responsible for the spatial distribution of \u003cem\u003eAphaenogaster\u003c/em\u003e-associated\u003cem\u003e\u0026nbsp;Wolbachia\u003c/em\u003e in GSMNP, then, in keeping with our findings, there should be fewer \u003cem\u003eWolbachia\u003c/em\u003e-positive ants on the eastern side of the park. However, \u003cem\u003eWolbachia\u003c/em\u003e-positiveants that \u003cem\u003edo\u003c/em\u003e reach the eastern side of GSMNP should have similar \u003cem\u003eWolbachia\u003c/em\u003e loads to their \u003cem\u003eWolbachia\u003c/em\u003e-positive counterparts on the western side of the park(i.e., the frequency of \u003cem\u003eWolbachia\u003c/em\u003e-positive ants should decrease, but the relative abundance of \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eon \u003cem\u003eWolbachia-\u003c/em\u003epositiveants should be similar everywhere). The fact that both prevalence and load decreased along the West-to-East gradient indicates that host dispersal is not the sole driver of \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003espatial patterns. That said, host-scale dispersal limitation does seem to exist, at least based on COI sequence (see SI II Fig. S2.5A) and thus, could partially explain the smaller fraction of \u003cem\u003eWolbachia\u003c/em\u003e-positive ants on the East side of the park - even if it doesn’t explain the overall lower \u003cem\u003eWolbachia\u003c/em\u003e loads in eastern ants when \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eis present.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eIndependent effects of microbe-scale processes\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe inferred independent effects of microbe-scale processes based on the significantly different\u0026nbsp;\u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eloads across host mitotypes, even in paired samples from the same sets of sites. Notably, this pattern could have emerged either through microbe-scale environmental selection or microbe-scale dispersal limitation. Microbe-scale environmental selection would indicate that the mitochondrial COI sequence is correlated (not necessarily causally) with some aspect of host biology (e.g., immune system, behavior, food preference) that impacts \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eperformance. Again, this would not be unprecedented. Among leaf-cutter ants (\u003cem\u003eAcromyrmex echinatior\u003c/em\u003e), for example, different patrilines exhibit significant differences in \u003cem\u003eWolbachia\u003c/em\u003e densities.\u003csup\u003e57\u003c/sup\u003e Meanwhile among \u003cem\u003eVollenhovia emeryi\u003c/em\u003e, genotypes associated with different wing morphology exhibit different \u003cem\u003eWolbachia\u003c/em\u003e colonization rates.\u003csup\u003e58\u003c/sup\u003e Even among \u003cem\u003eAphaenogaster\u003c/em\u003e, a recent study identified differences in \u003cem\u003eWolbachia\u003c/em\u003e abundances between \u003cem\u003eA. rudis\u003c/em\u003e and \u003cem\u003eA. fulva\u0026nbsp;\u003c/em\u003eants, although these results were based on only three \u003cem\u003eA. fulva\u003c/em\u003e and two \u003cem\u003eA. rudis\u003c/em\u003e colonies.\u003csup\u003e59\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eThe alternate explanation – that \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eare limited to red group ants because of microbe-scale dispersal limitation – would indicate a barrier to \u003cem\u003eWolbachia\u003c/em\u003e transmission among different host mitotypes. Again, this is consistent with what we know about \u003cem\u003eWolbachia\u003c/em\u003e biology.Broadly speaking, \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ehas two modes of microbe-scale dispersal: horizontal transmission and vertical transmission. Although extensive evidence exists suggesting that \u003cem\u003eWolbachia\u003c/em\u003e can be acquired horizontally across large host phylogenetic distances,\u003csup\u003e60\u003c/sup\u003e including among ants,\u003csup\u003e61\u003c/sup\u003e horizontal transmission within species requires a number of relatively restrictive conditions.\u003csup\u003e62\u003c/sup\u003e This likely limits the extent to which \u003cem\u003eWolbachia\u003c/em\u003e can transfer directly from one ant to another (microbe-scale dispersal limitation), particularly over shorter, ecological timescales. Consequently, in our system, microbe-scale dispersal of \u003cem\u003eWolbachia\u003c/em\u003e likely occurs predominantly via vertical transmission (mother to offspring).\u003csup\u003e18\u003c/sup\u003e As a result, \u003cem\u003eWolbachia\u003c/em\u003e colonization should closely mirror patterns in mtDNA inheritance, even if there is no correlation between mitochondrial COI sequence and other aspects of host biology impacting \u003cem\u003eWolbachia\u003c/em\u003e fitness (i.e., no microbe-scale environmental selection). Unfortunately, in our system, it is difficult to determine whether microbe-scale environmental selection, microbe-scale dispersal limitation or a combination of both underlies differences \u003cem\u003eWolbachia\u003c/em\u003e carriage across COI sequences. Nevertheless, the extremely tight coupling between COI sequence and each of the three \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eASVs, along with the limited number of \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eacquisitions and losses predicted by ancestral state reconstructions (ASRs, see SI III, Figures S.3.7 and S3.8), suggests that dispersal is strongly limited to vertical transmission. This makes microbe-scale dispersal limitation a strong contender for explaining the restriction of \u003cem\u003eWolbachia\u003c/em\u003e to red group ants and, to a lesser extent, purple clade ants. Consistent with this hypothesis, we did not find any relationships between other bacterial taxa (including both other endosymbionts and ectosymbionts), and COI sequence (see SI III, Figures S3.1 for visualizations), although it is possible that some of the other putative endosymbionts (e.g., \u003cem\u003eSulcia\u003c/em\u003e)were only detected because they were present in prey from \u003cem\u003eAphaenogaster\u0026nbsp;\u003c/em\u003estomach contents.\u003c/p\u003e\n\u003cp\u003eLike COI sequence, CAD sequence also appears to be correlated with some aspect of host biology that impacts \u003cem\u003eWolbachia\u003c/em\u003e performance. Notably, we found an effect of CAD sequence even within COI clades, suggesting that the correlation between CAD sequence and \u003cem\u003eWolbachia\u003c/em\u003e is not fully explained by the same drivers that underlie the correlation between COI sequence and \u003cem\u003eWolbachia\u003c/em\u003e. Likewise, CAD sequence remained correlated with \u003cem\u003eWolbachia\u003c/em\u003e load even when controlling for environment, as would be expected given that CAD is not, itself, correlated with environment. Unfortunately, the relatively few western locations harboring red clade ants, coupled with the low capture rate of red clade ants at eastern locations, made it difficult for us to test the effect of CAD within COI clades while simultaneously controlling for environment. Rather, we had only four sites for our paired Wilcoxon test, which did not, as a result, rise to the level of significance. Nevertheless, there was still evidence of a weak trend, suggesting that, with additional statistical power, it might be possible to detect independent effects of COI and CAD independent of environment (see Table 1). Regardless of whether COI and CAD sequences reflect the same or different aspects of host biology underlying \u003cem\u003eWolbachia\u003c/em\u003e load, there is no question that microbe-scale processes are an independent contributor to the spatial distribution of \u003cem\u003eAphaenogaster\u003c/em\u003e-associated \u003cem\u003eWolbachia\u003c/em\u003e across GSMNP.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePossible interactions between host and microbe scales\u003c/em\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eWhile we found conclusive evidence for independent effects of both host- and microbe-scale processes driving patterns in \u003cem\u003eAphaenogaster-\u003c/em\u003eassociated\u003cem\u003e\u0026nbsp;Wolbachia\u003c/em\u003e abundance across GSMNP, we also identified ample opportunity for multiscale interactions. Admittedly, such interactions are more difficult to demonstrate. Nevertheless, it is worth speculating on potential mechanisms that might connect host- and microbe-scale processes. Based on recent findings in \u003cem\u003eDrosophila\u003c/em\u003e, for example, it is possible that colder host environments (i.e., host-scale environmental selection) reduce vertical transmission (i.e., microbe-scale dispersal limitation) of \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003eto \u003cem\u003eAphaenogaster\u0026nbsp;\u003c/em\u003ealates.\u003csup\u003e63\u003c/sup\u003e This could explain the lower prevalence (though not necessarily the lower abundance) of \u003cem\u003eWolbachia\u0026nbsp;\u003c/em\u003ein ants, including red group ants, from the colder Eastern regions of the park. Another possible multiscale interaction is reproductive manipulation. If \u003cem\u003eWolbachia\u003c/em\u003e has been prevalent on the western side of the park for a long time, and if \u003cem\u003eWolbachia\u003c/em\u003e induces CI or some other form of reproductive manipulation in \u003cem\u003eA. rudis\u0026nbsp;\u003c/em\u003eants, then this could prevent western invasion of blue clade COI, even as blue clade nuclear genes (i.e., yellow clade CAD) become incorporated into red group populations (i.e., the microbe alters host-scale gene flow). Consistent with this hypothesis, we observed a higher percentage of red group COI (30%) mismatched with yellow clade CAD as compared to blue clade COI (21%) mismatched with green clade CAD. This is what would be expected if the offspring of uninfected females and infected males are non-viable but the offspring of infected females and uninfected males are viable. Further, while mismatch in blue clade COI is unexpected, this may be explained by mismatch that existed prior to \u003cem\u003eWolbachia\u003c/em\u003e colonization (akin to incomplete lineage sorting) or as a result of incomplete CI. Notably, temperature differences can result in differing levels of both CI penetrance and \u003cem\u003eWolbachia\u003c/em\u003e density,\u003csup\u003e55,64,65\u003c/sup\u003e providing a mechanism by which CI might serve as a stronger barrier to gene flow on the western side of the park, but a weaker barrier on the eastern side of the park. \u0026nbsp;Another potential multiscale mechanism that may contribute to our findings is an effect of \u003cem\u003eWolbachia\u003c/em\u003e on ant dispersal. This could occur either because \u003cem\u003eWolbachia\u003c/em\u003e alters the movement patterns of alates or because it reduces production of alates entirely.\u003csup\u003e66\u003c/sup\u003e In both cases, the effect would be to slow the eastern spread of red group COI, even if CI gives the red group a long-term competitive advantage.\u003csup\u003e67\u003c/sup\u003e Finally, \u003cem\u003eWolbachia\u003c/em\u003e could prevent red group ants from penetrating eastward by lowering survival in the colder, wetter, higher elevation regions of the park. Again, this multiscale interaction could restrict the spatial extent of the red group COI, even if it gains a competitive demographic advantage through CI.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eConclusions\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe complex genetic structure of \u003cem\u003eA. rudis\u0026nbsp;\u003c/em\u003eants, their propensity to inhabit a wide range of different environments across large spatial extents, and their colonization by multiple strains of the notoriously complex \u003cem\u003eWolbachia\u003c/em\u003e endosymbiont makes \u003cem\u003eA. rudis\u003c/em\u003e an ideal system for studying the interplay between environmental selection and dispersal limitation at the host and microbe scales. Our findings suggest that this interplay leads to interesting spatial patterns across the mountainous terrain in GSMNP. Further, multiple opportunities exist for extending the current study. This includes better characterizing \u003cem\u003eWolbachia\u003c/em\u003e differences in the contact zone between COI clades/groups, undertaking more extensive comparisons of \u003cem\u003eWolbachia\u003c/em\u003e carriage in ants with and without COI/CAD mismatch and extending similar analyses from GSMNP to the Southern Appalachians more broadly in order to better tease apart the effects of environment versus distance in shaping patterns of \u003cem\u003eWolbachia\u003c/em\u003e prevalence and abundance. Beyond field characterization, another fruitful avenue would be lab studies examining how \u003cem\u003eWolbachia\u003c/em\u003e impacts host thermal tolerance, how temperature effects \u003cem\u003eWolbachia\u003c/em\u003e load and transmission, or even whether and how \u003cem\u003eWolbachia\u003c/em\u003e induces CI or other forms of reproductive manipulation.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAs keystone mutualists, \u003cem\u003eAphaenogaster\u003c/em\u003e ants have garnered significant attention across various research areas,\u003csup\u003e27,29,33,42,68–72\u003c/sup\u003e including recent explorations into their microbial associations.\u003csup\u003e59,73,74\u003c/sup\u003e Our study suggests that, even beyond their importance to processes like seed dispersal and pest control,\u003csup\u003e75\u003c/sup\u003e\u003cem\u003eA. rudis\u0026nbsp;\u003c/em\u003ecomplex antsmight serve as model systems for understanding the multiscale, bidirectional interactions between hosts and their HA microbes.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003cb\u003eFunding Declaration\u003c/b\u003e: This research was made possible, in part, with support from the Clemson University Genomics and Bioinformatics Facility, which receives support from the College of Science and two Institutional Development Awards (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health under grant numbers P20GM146584 and P20GM139769.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eD.A.M., B.C. and S.B. designed the experiment and performed the fieldwork. D.A.M., C.H., E.R.G. and M.C. designed and performed sequencing of the host. D.A.M., B.C., S.B., M.A.S., S.M. and A.S. designed and performed sequencing of the microbiota. D.A.M., S.A.B., S.R.B. and S.R.B. designed and performed Wolbachia analysis. All authors contributed to data analysis and writing of the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe thank Dr. Megan Pitz for her help with creation of Figure 3 and her willingness to tolerate centipedes in her closet. We would also like to thank Nico Malagon, Dr. Tatsuhiro Kato, Daniel Cross, Jackson Lloyd, Drew Kanes, Simon Dunn, and Mr. Paul Super for their help in the collection of ant samples.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eData Accessibility and Benefit-Sharing:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eData Accessibility\u003c/em\u003e:The datasets generated and/or analysed during the current study are available in the National Center for Biotechnology Information (NCBI) sequence read archive (SRA) under project number PRJNA1347072. All data and code (BIOM files, metadata files, and all R code) necessary for the analyses presented in this manuscript are additionally available at https://github.com/dmalago/Apheanogaster_GSMNP_Microbiome.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eBenefit-Sharing\u003c/em\u003e: Research results have been shared with Great Smoky Mountains National Park through yearly reports.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eBellier, E. et al. Distance decay of similarity, effects of environmental noise and ecological heterogeneity among species in the spatio-temporal dynamics of a dispersal‐limited community. \u003cem\u003eEcography\u003c/em\u003e \u003cb\u003e37\u003c/b\u003e, 172\u0026ndash;182 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGl\u0026uuml;ck, M., Geue, J. C. \u0026amp; Thomassen, H. A. Environmental differences explain subtle yet detectable genetic structure in a widespread pollinator. \u003cem\u003eBMC Ecol. Evol.\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e, 8 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcRae, B. H. ISOLATION BY RESISTANCE. \u003cem\u003eEvol. (N Y)\u003c/em\u003e. \u003cb\u003e60\u003c/b\u003e, 1551\u0026ndash;1561 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoininen, J., McDonald, R. \u0026amp; Hillebrand, H. The distance decay of similarity in ecological communities. \u003cem\u003eEcography\u003c/em\u003e \u003cb\u003e30\u003c/b\u003e, 3\u0026ndash;12 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBaguette, M., Blanchet, S., Legrand, D., Stevens, V. M. \u0026amp; Turlure, C. Individual dispersal, landscape connectivity and ecological networks. \u003cem\u003eBiol. Rev.\u003c/em\u003e \u003cb\u003e88\u003c/b\u003e, 310\u0026ndash;326 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCouch, C. E., Epps, C. W. \u0026amp; Host Microbiome, and Complex Space: Applying Population and Landscape Genetic Approaches to Gut Microbiome Research in Wild Populations. \u003cem\u003eJ. Hered.\u003c/em\u003e \u003cb\u003e113\u003c/b\u003e, 221\u0026ndash;234 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRuuskanen, M. O., Sommeria-Klein, G., Havulinna, A. S., Niiranen, T. J. \u0026amp; Lahti, L. Modelling spatial patterns in host‐associated microbial communities. \u003cem\u003eEnviron. Microbiol.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e, 2374\u0026ndash;2388 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrsini, L., Vanoverbeke, J., Swillen, I., Mergeay, J. \u0026amp; De Meester, L. Drivers of population genetic differentiation in the wild: isolation by dispersal limitation, isolation by adaptation and isolation by colonization. \u003cem\u003eMol. Ecol.\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e, 5983\u0026ndash;5999 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHutchison, D. W., Templeton, A. R., CORRELATION OF PAIRWISE GENETIC \u0026amp; AND GEOGRAPHIC DISTANCE MEASURES: INFERRING THE RELATIVE INFLUENCES OF GENE FLOW AND DRIFT ON THE DISTRIBUTION OF GENETIC VARIABILITY. \u003cem\u003eEvol. (N Y)\u003c/em\u003e \u003cb\u003e53\u003c/b\u003e, 1898\u0026ndash;1914 (1999).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWright, S. \u003cem\u003eISOLATION DISTANCE Genetics\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e, 114\u0026ndash;138 (1943).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, G. et al. Host-microbiota interaction helps to explain the bottom-up effects of climate change on a small rodent species. \u003cem\u003eISME J.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 1795\u0026ndash;1808 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRisely, A., Klaassen, M. \u0026amp; Hoye, B. J. Migratory animals feel the cost of getting sick: A meta-analysis across species. \u003cem\u003eJ. Anim. Ecol.\u003c/em\u003e \u003cb\u003e87\u003c/b\u003e, 301\u0026ndash;314 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndriolli, F. S. et al. Do zombie ant fungi turn their hosts into light seekers? \u003cem\u003eBehav. Ecol.\u003c/em\u003e \u003cb\u003e30\u003c/b\u003e, 609\u0026ndash;616 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVyas, A., Kim, S. K., Giacomini, N., Boothroyd, J. C. \u0026amp; Sapolsky, R. M. Behavioral changes induced by Toxoplasma infection of rodents are highly specific to aversion of cat odors. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 104, 6442\u0026ndash;6447 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLemoine, M. M., Engl, T. \u0026amp; Kaltenpoth, M. Microbial symbionts expanding or constraining abiotic niche space in insects. \u003cem\u003eCurr. Opin. Insect Sci.\u003c/em\u003e \u003cb\u003e39\u003c/b\u003e, 14\u0026ndash;20 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOtti, O. Genitalia-associated microbes in insects. \u003cem\u003eInsect Sci.\u003c/em\u003e \u003cb\u003e22\u003c/b\u003e, 325\u0026ndash;339 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTelschow, A., Hammerstein, P. \u0026amp; Werren, J. H. The Effect of Wolbachia on Genetic Divergence between Populations: Models with Two-Way Migration. \u003cem\u003eAm. Nat.\u003c/em\u003e \u003cb\u003e160\u003c/b\u003e, S54\u0026ndash;S66 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaur, R. et al. Living in the endosymbiotic world of Wolbachia: A centennial review. \u003cem\u003eCell. Host Microbe\u003c/em\u003e. \u003cb\u003e29\u003c/b\u003e, 879\u0026ndash;893 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeeks, A. R., Reynolds, T., Hoffmann, A. A. \u0026amp; K. \u0026amp; Wolbachia dynamics and host effects: what has (and has not) been demonstrated? \u003cem\u003eTrends Ecol. Evol.\u003c/em\u003e \u003cb\u003e17\u003c/b\u003e, 257\u0026ndash;262 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeinert, L. A., Araujo-Jnr, E. V., Ahmed, M. Z. \u0026amp; Welch, J. J. The incidence of bacterial endosymbionts in terrestrial arthropods. \u003cem\u003eProceedings of the Royal Society B: Biological Sciences\u003c/em\u003e 282, 20150249 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCalvitti, M., Marini, F., Desiderio, A., Puggioli, A. \u0026amp; Moretti, R. Wolbachia Density and Cytoplasmic Incompatibility in Aedes albopictus: Concerns with Using Artificial Wolbachia Infection as a Vector Suppression Tool. \u003cem\u003ePLoS One\u003c/em\u003e. \u003cb\u003e10\u003c/b\u003e, e0121813 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTortosa, P. et al. Wolbachia Age-Sex-Specific Density in Aedes albopictus: A Host Evolutionary Response to Cytoplasmic Incompatibility? \u003cem\u003ePLoS One\u003c/em\u003e. \u003cb\u003e5\u003c/b\u003e, e9700 (2010).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaenike, J., Dyer, K. A., Cornish, C. \u0026amp; Minhas, M. S. Asymmetrical Reinforcement and Wolbachia Infection in Drosophila. \u003cem\u003ePLoS Biol.\u003c/em\u003e \u003cb\u003e4\u003c/b\u003e, e325 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCruz, M. A., Magalh\u0026atilde;es, S., Sucena, \u0026Eacute;. \u0026amp; Z\u0026eacute;l\u0026eacute;, F. Wolbachia and host intrinsic reproductive barriers contribute additively to postmating isolation in spider mites. \u003cem\u003eEvol. (N Y)\u003c/em\u003e. \u003cb\u003e75\u003c/b\u003e, 2085\u0026ndash;2101 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFridley, J. D. Downscaling Climate over Complex Terrain: High Finescale (\u0026lt;1000 m) Spatial Variation of Near-Ground Temperatures in a Montane Forested Landscape (Great Smoky Mountains)*. \u003cem\u003eJ Appl Meteorol Climatol\u003c/em\u003e 48, 1033\u0026ndash;1049 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStark, J. R. \u0026amp; Fridley, J. D. Topographic Drivers of Soil Moisture Across a Large Sensor Network in the Southern Appalachian Mountains (USA). \u003cem\u003eWater Resour. Res\u003c/em\u003e \u003cb\u003e59\u003c/b\u003e, (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLubertazzi, D. The biology and natural history of Aphaenogaster rudis. \u003cem\u003ePsyche (Stuttg)\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1155/2012/752815\u003c/span\u003e\u003cspan address=\"10.1155/2012/752815\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaysen, E. E. S. \u0026amp; DIVERSITY AND ABUNDANCE OF ANTS AT FOREST EDGES IN THE GREAT SMOKY MOUNTAINS NATIONAL PARK. undefined-undefined (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHelms Cahan, S. et al. Modulation of the heat shock response is associated with acclimation to novel temperatures but not adaptation to climatic variation in the ants Aphaenogaster picea and A. rudis. \u003cem\u003eComp. Biochem. Physiol. Mol. Integr. Physiol.\u003c/em\u003e \u003cb\u003e204\u003c/b\u003e, 113\u0026ndash;120 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLessard, J. P., Dunn, R. R., Parker, C. R. \u0026amp; Sanders, N. J. Rarity and Diversity in Forest Ant Assemblages of Great Smoky Mountains National Park. \u003cem\u003eSoutheast. Nat.\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, 215\u0026ndash;228 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLynch, J. F. \u0026amp; Seasonal Successional, and Vertical Segregation in a Maryland Ant Community. \u003cem\u003eOikos\u003c/em\u003e \u003cb\u003e37\u003c/b\u003e, 183 (1981).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFellers, J. H. Interference and exploitation in a guild of woodland ants. \u003cem\u003eEcology\u003c/em\u003e \u003cb\u003e68\u003c/b\u003e, 1466\u0026ndash;1478 (1987).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBanschbach, V. S., Brunelle, A., Bartlett, K. M., Grivetti, J. Y. \u0026amp; Yeamans, R. L. Tool use by the forest ant Aphaenogaster rudis: Ecology and task allocation. \u003cem\u003eInsectes Soc.\u003c/em\u003e \u003cb\u003e53\u003c/b\u003e, 463\u0026ndash;471 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCole, A. C. A Guide to the Ants of the Great Smoky Mountains National Park, Tennessee. \u003cem\u003eAm. Midl. Nat.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e, 1\u0026ndash;88 (1940).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWarren, R. J. et al. Climate-driven range shift prompts species replacement. \u003cem\u003eInsectes Soc.\u003c/em\u003e \u003cb\u003e63\u003c/b\u003e, 593\u0026ndash;601 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWarren, R. J., McAfee, P. \u0026amp; Bahn, V. Ecological differentiation among key plant mutualists from a cryptic ant guild. \u003cem\u003eInsectes Soc.\u003c/em\u003e \u003cb\u003e58\u003c/b\u003e, 505\u0026ndash;512 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBuono, C. M. et al. Uncovering how behavioral variation underlying mutualist partner quality is partitioned within a species complex of keystone seed-dispersing ants. \u003cem\u003eInsectes Soc.\u003c/em\u003e \u003cb\u003e69\u003c/b\u003e, 247\u0026ndash;260 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNichols, B. J. \u0026amp; Langdon, K. R. The Smokies All Taxa Biodiversity Inventory: History and Progress. \u003cem\u003eSoutheastern Naturalist\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCaterino, M. S. \u0026amp; Harden, C. W. Unseeing and Unseen: On the Distribution, Morphology, and Larva of One of North America\u0026rsquo;s Rarest Histerid Beetles, Geocolus caecus Wenzel (Coleoptera: Histeridae). \u003cem\u003eColeopt Bull\u003c/em\u003e \u003cb\u003e76\u003c/b\u003e, (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTeam, R. C. \u0026amp; R Core Team. R Core Team. \u0026amp; R: A language and environment for statistical computing. \u003cem\u003eR Foundation for Statistical Computing, Vienna, Austria. URL\u003c/em\u003e (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.R-project.org/.\u003c/span\u003e\u003cspan address=\"http://www.R-project.org/.\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e R Foundation for Statistical Computing (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChao, K. H., Barton, K., Palmer, S. \u0026amp; Lanfear, R. sangeranalyseR: Simple and Interactive Processing of Sanger Sequencing Data in R. \u003cem\u003eGenome Biol. Evol\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeMarco, B. B. \u0026amp; Cognato, A. I. A multiple-gene phylogeny reveals polyphyly among eastern North American Aphaenogaster species (Hymenoptera: Formicidae). \u003cem\u003eZool. Scr.\u003c/em\u003e \u003cb\u003e45\u003c/b\u003e, 512\u0026ndash;520 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTrifinopoulos, J., Nguyen, L. T., von Haeseler, A. \u0026amp; Minh, B. Q. W-IQ-TREE: a fast online phylogenetic tool for maximum likelihood analysis. \u003cem\u003eNucleic Acids Res.\u003c/em\u003e \u003cb\u003e44\u003c/b\u003e, W232\u0026ndash;W235 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLetten, A. D. \u0026amp; Cornwell, W. K. Trees, branches and (square) roots: why evolutionary relatedness is not linearly related to functional distance. \u003cem\u003eMethods Ecol. Evol.\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e, 439\u0026ndash;444 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKozich, J. J., Westcott, S. L., Baxter, N. T., Highlander, S. K. \u0026amp; Schloss, P. D. Development of a Dual-Index Sequencing Strategy and Curation Pipeline for Analyzing Amplicon Sequence Data on the MiSeq Illumina Sequencing Platform. \u003cem\u003eAppl. Environ. Microbiol.\u003c/em\u003e \u003cb\u003e79\u003c/b\u003e, 5112\u0026ndash;5120 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBolyen, E. et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. \u003cem\u003eNat. Biotechnol.\u003c/em\u003e \u003cb\u003e37\u003c/b\u003e, 852\u0026ndash;857 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNelder, J. A. \u0026amp; Wedderburn, R. W. M. Generalized Linear Models. \u003cem\u003eJ. R Stat. Soc. Ser. A\u003c/em\u003e. \u003cb\u003e135\u003c/b\u003e, 370 (1972).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMantel, N. The detection of disease clustering and a generalized regression approach. \u003cem\u003eCancer Res\u003c/em\u003e (1967).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJari Oksanen, R. et al. Package \u0026lsquo;vegan\u0026rsquo;. (2007). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.researchgate.net/publication/228339454_The_vegan_Package?enrichId=rgreq-87dd17328dd562faca4cd6cd67fd8d2b-XXX\u0026amp;enrichSource=Y292ZXJQYWdlOzIyODMzOTQ1NDtBUzoxMDIxMjA\u003c/span\u003e\u003cspan address=\"https://www.researchgate.net/publication/228339454_The_vegan_Package?enrichId=rgreq-87dd17328dd562faca4cd6cd67fd8d2b-XXX\u0026amp;enrichSource=Y292ZXJQYWdlOzIyODMzOTQ1NDtBUzoxMDIxMjA\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cem\u003e3NjYzMTI0NjVAMTQwMTM1ODg5NzY3NQ%3D%3D⪙=1_x_2\u0026amp;_esc=publicationCoverPdf\u003c/em\u003e http://cran.ism.ac.jp/web/packages/vegan/vegan.pdf\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParadis, E. \u0026amp; Schliep, K. ape 5.0: an environment for modern phylogenetics and evolutionary analyses in {R}. \u003cem\u003eBioinformatics\u003c/em\u003e \u003cb\u003e35\u003c/b\u003e, 526\u0026ndash;528 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConover, W. J. \u003cem\u003ePractical Nonparametric Statistics\u003c/em\u003e. (1999).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMOUTON, L., HENRI, H., BOULETREAU, M. \u0026amp; VAVRE, F. Effect of temperature on Wolbachia density and impact on cytoplasmic incompatibility. \u003cem\u003eParasitology\u003c/em\u003e \u003cb\u003e132\u003c/b\u003e, 49\u0026ndash;56 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHurst, G. D. D., Jiggins, F. M. \u0026amp; Robinson, S. J. W. What causes inefficient transmission of male-killing Wolbachia in Drosophila? \u003cem\u003eHeredity (Edinb)\u003c/em\u003e. \u003cb\u003e87\u003c/b\u003e, 220\u0026ndash;226 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKelleher, L. A. \u0026amp; Ramalho, M. O. Impact of rising temperatures on the bacterial communities of Aphaenogaster ants. \u003cem\u003eBiol Open\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBordenstein, S. R. \u0026amp; Bordenstein, S. R. Temperature Affects the Tripartite Interactions between Bacteriophage WO, Wolbachia, and Cytoplasmic Incompatibility. \u003cem\u003ePLoS One\u003c/em\u003e. \u003cb\u003e6\u003c/b\u003e, e29106 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCharlesworth, J., Weinert, L. A., Araujo, E. V., Welch, J. J. \u0026amp; Wolbachia Cardinium and climate: an analysis of global data. \u003cem\u003eBiol. Lett.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 20190273 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFrost, C. L., Mitchell, R., Smith, J. E. \u0026amp; Hughes, W. O. H. Genotypes and phenotypes in a Wolbachia -ant symbiosis. \u003cem\u003ePeerJ\u003c/em\u003e 12, e17781 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoh, P. et al. Association between host wing morphology polymorphism and Wolbachia infection in Vollenhovia emeryi (Hymenoptera: Myrmicinae). \u003cem\u003eEcol. Evol.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 8827\u0026ndash;8837 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKelleher, L. A. \u0026amp; Ramalho, M. O. Impact of Species and Developmental Stage on the Bacterial Communities of Aphaenogaster Ants. \u003cem\u003eCurr. Microbiol.\u003c/em\u003e \u003cb\u003e82\u003c/b\u003e, 157 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhmed, M. Z., Breinholt, J. W. \u0026amp; Kawahara, A. Y. Evidence for common horizontal transmission of Wolbachia among butterflies and moths. \u003cem\u003eBMC Evol. Biol.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e, 118 (2016).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRamalho, M., de Kim, O., Wang, Z., Moreau, C. S. \u0026amp; S. \u0026amp; Wolbachia Across Social Insects: Patterns and Implications. \u003cem\u003eAnn. Entomol. Soc. Am.\u003c/em\u003e \u003cb\u003e114\u003c/b\u003e, 206\u0026ndash;218 (2021).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTolley, S. J. A., Nonacs, P. \u0026amp; Sapountzis, P. Wolbachia Horizontal Transmission Events in Ants: What Do We Know and What Can We Learn? \u003cem\u003eFront Microbiol\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHague, M. T. J. et al. Temperature effects on cellular host-microbe interactions explain continent-wide endosymbiont prevalence. \u003cem\u003eCurr. Biol.\u003c/em\u003e \u003cb\u003e32\u003c/b\u003e, 878\u0026ndash;888e8 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHurst, G. D. D., Johnson, A. P., Schulenburg, J. H. \u0026amp; Fuyama, Y. G. v d Male-Killing Wolbachia in Drosophila: A Temperature-Sensitive Trait With a Threshold Bacterial Density. \u003cem\u003eGenetics\u003c/em\u003e 156, 699\u0026ndash;709 (2000).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMOUTON, L., HENRI, H., BOULETREAU, M. \u0026amp; VAVRE, F. Effect of temperature on Wolbachia density and impact on cytoplasmic incompatibility. \u003cem\u003eParasitology\u003c/em\u003e \u003cb\u003e132\u003c/b\u003e, 49\u0026ndash;56 (2006).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWenseleers, T., Sundstr\u0026ouml;m, L. \u0026amp; Billen, J. Deleterious Wolbachia in the ant Formica truncorum. \u003cem\u003eProc. R Soc. Lond. B Biol. Sci.\u003c/em\u003e \u003cb\u003e269\u003c/b\u003e, 623\u0026ndash;629 (2002).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCariou, M., Duret, L. \u0026amp; Charlat, S. The global impact of Wolbachia on mitochondrial diversity and evolution. \u003cem\u003eJ. Evol. Biol.\u003c/em\u003e \u003cb\u003e30\u003c/b\u003e, 2204\u0026ndash;2210 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNelson, B. Y., INCREASED FUNGAL DIVERSITY ASSOCIATED \u0026amp; WITH APHAENOGASTER SPP. \u003cem\u003eMORE EVIDENCE FOR KEYSTONE MUTUALISMS\u003c/em\u003e (Western Carolina University, 2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUniversity, M. N. S. W. C. \u0026amp; undefined Effects of myrmecochore species abundance, diversity, and fruiting phenology on Aphaenogaster (Hymenoptera: Formicidae) nesting and foraging in Southern. \u003cem\u003ecore.ac.ukMN SchultzWestern Carolina University, 2014\u0026bull;core.ac.uk\u003c/em\u003e (2014). (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNess, J. H., Morin, D. F. \u0026amp; Giladi, I. Uncommon specialization in a mutualism between a temperate herbaceous plant guild and an ant: Are Aphaenogaster ants keystone mutualists? \u003cem\u003eOikos\u003c/em\u003e 118, 1793\u0026ndash;1804 (2009).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUmphrey, G. J. Morphometric discrimination among sibling species in the fulva \u0026ndash; rudis \u0026ndash; texana complex of the ant genus Aphaenogaster (Hymenoptera: Formicidae). \u003cem\u003eCan. J. Zool.\u003c/em\u003e \u003cb\u003e74\u003c/b\u003e, 528\u0026ndash;559 (1996).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaysen, E. S. et al. FORMICIDAE) NESTING AND FORAGING IN SOUTHERN APPALACHIAN RICH COVE FORESTS. \u003cem\u003eEcology\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 5\u0026ndash;7 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLenoir, A. \u0026amp; Devers, S. Alkaloid secretion inhibited by antibiotics in Aphaenogaster ants. \u003cem\u003eC R Biol.\u003c/em\u003e \u003cb\u003e341\u003c/b\u003e, 358\u0026ndash;361 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePagalilauan, A., Pavloudi, C., Ospina, S. M., Smith, A. \u0026amp; Saw, J. Interaction with refuse piles is associated with co-occurrence of core gut microbiota in workers of the ant Aphaenogaster picea. \u003cem\u003ePreprint at.\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1099/acmi.0.000832.v3\u003c/span\u003e\u003cspan address=\"10.1099/acmi.0.000832.v3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerfecto, I., Philpott, S. M. \u0026amp; Ants (Hymenoptera: Formicidae) and ecosystem functions and services in urban areas: a reflection on a diverse literature. \u003cem\u003eMyrmecol News\u003c/em\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1.\u003c/strong\u003e Results of paired sample Wilcox tests comparing average relative abundance of \u003cem\u003eWolbachia\u003c/em\u003e from ants where red, purple and/or blue clade ants co-occur.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"485\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 250px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eComparison\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 113px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eV-statistic\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003e\u003cstrong\u003ep-value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 250px;\"\u003e\n \u003cp\u003eblue:red\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 113px;\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003e0.0625 .\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 250px;\"\u003e\n \u003cp\u003eblue:purple\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 113px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003e0.125\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 250px;\"\u003e\n \u003cp\u003epurple:red\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 113px;\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003e0.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 250px;\"\u003e\n \u003cp\u003ered+purple:blue\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 113px;\"\u003e\n \u003cp\u003e28 \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003e0.007813**\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 250px;\"\u003e\n \u003cp\u003eyellow:green\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 113px;\"\u003e\n \u003cp\u003e37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003e0.049*\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"bottom\" style=\"width: 250px;\"\u003e\n \u003cp\u003ered+purple/yellow:red+purple/green\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 113px;\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"bottom\" style=\"width: 122px;\"\u003e\n \u003cp\u003e0.125\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Aphaenogaster, Wolbachia, host-associated microbes, multiscale, environmental selection, dispersal limitation","lastPublishedDoi":"10.21203/rs.3.rs-8515672/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8515672/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe spatial distributions of host-associated (HA) microbes are shaped by the spatial processes of environmental selection and dispersal. However, unlike free-living organisms, HA microbes experience selection and dispersal at two separate spatial scales \u0026ndash; the scale of the microbes and the scale of their hosts. Therefore, HA microbes must tolerate both the environment created by their host (microbe-scale environment) and the environment in which their host resides (host-scale environment). Likewise, HA microbes can disperse both between hosts through horizontal or vertical transmission (microbe-scale dispersal) and between locations through host movement (host-scale dispersal). In this paper, we examine how host- and microbe-scale spatial processes contribute to the spatial distribution of \u003cem\u003eWolbachia\u003c/em\u003e endosymbionts in \u003cem\u003eAphaenogaster fulva-rudis-texana\u003c/em\u003e (Hymenoptera: Formicidae) complex ants from Great Smoky Mountains National Park. We begin by identifying significant spatial variation in \u003cem\u003eWolbachia\u003c/em\u003e relative abundance at both the host (across the landscape) and microbe (across host lineages) scales. We then demonstrate a correlation between host- and microbe-scale environmental selection, complicating efforts to isolate the independent effects of host- versus microbe-scale processes. To overcome this challenge, we leverage both the broad distributions of individual host lineages across different environments and sites of co-occurrence between different host lineages within the same environments. This allows us to assess how both host- and microbe-scale processes contribute to spatial variation in our system. Ultimately, our results shed light on the myriad of interacting factors governing spatial variation in HA microbes and why spatial variation in HA microbes is more challenging to understand than spatial variation in free-living organisms.\u003c/p\u003e","manuscriptTitle":"Host and Microbe Scale Processes Jointly Shape Spatial Variation in Aphaenogaster (Hymenoptera: Formicidae)associated Wolbachia","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-29 10:11:55","doi":"10.21203/rs.3.rs-8515672/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-03-02T06:26:44+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-17T15:49:43+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-16T01:00:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"66385220863815746558656402519173952284","date":"2026-02-04T00:03:10+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"249645165679786370516364749479713682364","date":"2026-02-02T20:00:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"66385220863815746558656402519173952284","date":"2026-02-01T17:28:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"296209784668545673616765511249233811713","date":"2026-01-29T20:40:48+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"10267413633063008977021519920776311913","date":"2026-01-29T14:16:48+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-01-27T07:28:14+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-01-07T13:17:40+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-05T06:55:55+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-05T06:54:38+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-01-04T23:13:08+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"1da95596-e23c-413f-b10b-8832fa1dcf25","owner":[],"postedDate":"January 29th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":61896911,"name":"Biological sciences/Ecology"},{"id":61896912,"name":"Earth and environmental sciences/Ecology"},{"id":61896913,"name":"Biological sciences/Microbiology"}],"tags":[],"updatedAt":"2026-05-15T04:09:31+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-29 10:11:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8515672","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8515672","identity":"rs-8515672","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}