Introduction

Coral reefs are complex ecosystems that rely on the symbiotic relationship between Scleractinia coral and the dinoflagellate Symbiodiniaceae (Falkowski et al. 1984; LaJeunesse et al. 2018; Muscatine et al. 1981). The symbiotic algae are photoautotrophs that convert sunlight and carbon dioxide into organic carbon and oxygen (Muscatine and Porter 1977; Page et al. 2019). However, rising sea surface temperatures (SST) have led to more frequent and severe coral bleaching events, resulting in the loss of their symbionts and subsequent mortality of the coral host (Baker 2003; Glynn 1984; Heron et al. 2016; Hoegh-Guldberg 1999; Hughes et al. 2003, 2017; Rowan et al. 1997). Understanding the effects of extreme environmental conditions on reef-building corals is crucial for predicting the potential impact of future ocean conditions on coral populations (Camp et al. 2018; Kleypas et al. 1999).

The diversity in both genetic and functional aspects of algal symbionts within the Symbiodiniaceae family is extensive (LaJeunesse et al. 2018) and particular coral-algae partnerships have been suggested to enhance the resilience of the coral holobiont to climate change (Berkelmans and Van Oppen 2006; Howells et al. 2011). Recent studies indicate that different genera of Symbiodiniaceae form symbiotic relationships with various coral hosts and exhibit distinct physiological attributes (e.g. enhanced growth rates, photosynthesis efficiency, thermal tolerance) across extensive latitudinal and environmental gradients (D. M. Baker et al. 2013; Chen et al. 2019; Hume et al. 2016; LaJeunesse et al. 2010, 2018; Ziegler et al. 2015). The performance and functionality of the coral holobiont is therefore, influenced by the identity and composition of its symbionts' distinct genetics and physiology (Johnston et al. 2022; Lajeunesse et al. 2014, 2018).

Investigating corals in various types of habitats such as turbid or high-latitude reefs offers a unique opportunity to uncover the mechanisms that support the ecological resilience of the coral communities persisting at extreme environmental conditions (Smith et al. 2020) Environmental factors, such as irradiance (Rowan et al. 1997), turbidity (Garren et al. 2006) and temperature (LaJeunesse et al. 2010; Tonk et al. 2013) have been linked to diversity and specificity of the coral-algal symbiosis. In extreme environments, corals hosts have been found to associate more with symbiont communities characterized by the stress tolerant Durusdinium, novel Cladocopium clades, and/or high symbiont diversity (Hennige et al. 2010; Oliver and Palumbi 2011; Smith et al. 2017; Wicks et al. 2010). Enhancing our knowledge of symbiont community structure and specific site/host associations is central to understanding how these relationships contribute to coral resilience in the face of climate change (Baker 2001; LaJeunesse et al. 2004).

The coral reefs of Western Australia (WA) span a wide latitudinal and environmental range, from tropical regions at 13̊ S to temperate waters at 35̊ S, where coral reefs owe their existence to the influential Leeuwin current, which transports warm water along the continental shelf (Cresswell and Golding 1980). However, due to the widespread and isolated nature of WA coral reefs, limited research has been conducted on the symbiotic relationships between coral hosts and their symbiont algae. As a result, current knowledge of the biogeographic patterns of coral symbiont communities across large latitudinal gradients is limited (Silverstein et al. 2011; Thomas et al. 2014, 2019).

Population genetic studies in reef coral from WA suggested that the high-latitude temperate reefs are isolated from potential recruits originating in tropical waters (Evans et al. 2019, 2021; Thomas et al. 2017). In particular, Evans et al., (2021) study on Turbinaria reniformis across a tropical-temperate transition zone, found pronounced regional genetic differentiation and isolation of high-latitude coral populations. T. reniformis, a broadcast- spawning coral species, is considered a generalist species distributed throughout the Indo-West pacific, ranging from southern to northern WA (Veron 1993). Since, rising sea temperatures threaten the future of reefs globally (Hoegh-Guldberg et al. 2007), knowledge on T. reniformis and its associated symbionts is crucial for understanding the species’ widespread geographical distribution and its ability to adapt to varying environmental drivers.

Here, we build upon the work of Evans et al. (2021) by examining the symbiont communities using metabarcoding of the internal transcribed spacer region 2 (ITS2) in T. reniformis collected from six sites along the WA coast, spanning 12 degrees of latitude and a transition from tropical to temperate waters. We hypothesised, that similar to the coral host, the symbiont community associated with T. reniformis would display strong structure and a shift in community composition along the latitudinal gradient due to the different environmental conditions and large distances between sample sites. Our main objectives were to (1) explore patterns of symbiont community composition and diversity in a single host species across a latitudinal and environmental gradient, (2) examine the effect of environmental drivers on the symbiont community assemblages, and (3) determine whether these patterns are associated with the host genetic structure.

Methods and materials

Study sites

Sampling was conducted at six sites (Evans et al. 2021), situated in three geographical areas (Fig. 1a). The four most northern sites include Balla Balla (BB: 20°38′15″S, 117°39′42″E), Dampier Archipelago (DA: 20°32′12″S, 116°46′37″E), Montebello Islands (MI: 20°31′16″S, 115°32′39″E), and Exmouth Gulf (EG: 22°17′41″S, 114°8′59″E). All falling within a similar biogeographic region known as the Pilbara, these reefs are a renowned biodiversity hotspot, particularly for fish populations (McLean et al. 2016) and filter-feeding communities (Fromont et al. 2016). Within this region, coral reefs are found along a semi-arid coastline with a natural cross-shelf turbidity gradient (Evans et al. 2020; Moustaka et al. 2018), and low light penetration at nearshore sites (Abdul Wahab et al. 2017). Additionally, the coastal waters of the Pilbara are impacted by tropical cyclones (Lough 1998) which increase turbidity and can cause physical damage to corals (Harmelin-Vivien 1994).

Fig. 1
figure 1

a Map of sampling sites along Western Australia coast. b Distance based dendrogram of 7,117 SNPs representing the relationship among all Turbinaria spp. samples collected by Evans et al 2021. Samples are colour coded by COI haplotype. Orange dots highlight the single SNP lineage and mitochondrial haplotype (H2) group that was determined to be T. reniformis using museum voucher specimens and was the focus of this study. c Population genetic structure of T. reniformis (n = 55; H2 haplotypes) host samples used in this study represented by a scatter plot of discriminant analysis of principal components (DAPC)

Full size image

Moving south, Shark Bay (SB, 25°51′5″S 113°20′52″E), is a UNESCO World Heritage site, located in the sub-tropical Gascoyne region, central Western Australia. In the bay, high salinities and steep temperature gradients (19–27 °C) occur due to its shallow depths (< 20 m) and presence of barrier islands. As such, coral growth is restricted to the seaward margins of the inner embayments, which are more frequently exposed to oceanic water (Marsh 1990; O’Leary et al. 2008). Shark Bay is a transition area between temperate and tropical marine zones. As such, it encompasses a mixture of temperate and tropical species as well as endemic species to Western Australia (Wyatt et al. 2005).

The most southern site, Perth (Pe, 31°41′33″S 115°38′58″E), represents a high-latitude, offshore coral reef located in the Marmion Marine Park approximately 40 km from Perth city. Here, complex chains of reefs, coves and lagoons act as barriers, restricting water exchange between inshore and offshore waters (Pobar et al. 1992). The dominant climatic factor here is wind, which generates waves, induces water circulation and transports sand inland (Pobar et al. 1992).

Sample collection and DNA extractions

In this study we focussed on the turbid reef specialist, Turbinaria reniformis. We used samples from a recent population genetic study that explored T. reniformis host structure across a latitudinal gradient in Western Australia (Evans et al. 2021). In their study, Evans et al., (2021), identified strong geographic structure in the coral host along the coastline of Western Australia based on a panel of thousands of single nucleotide polymorphism (Fig. 1b, c). We used these samples to explore patterns of symbiont community composition and diversity, allowing us to compare host patterns observed with symbiont communities, and to provide assurances that we were focussed on a single species and gene pool for our analyses. Briefly, samples of 2 cm2 nubbins were collected on SCUBA using side cutters, from colonies at least 5 m apart to avoid clone mates but within 100 m2. Samples were preserved in 100% AR grade ethanol. DNA was extracted and purified using a Qiagen Dneasy Blood & Tissue kit (plate format) following the manufacturer’s protocol and stored in − 80 °C until further processing.

Environmental data

To detect any differences in environmental drivers among sites, daily composite data of diffused attenuation (Kd490, used as a proxy for turbidity), chlorophyll a (Chl a), and sea surface temperature (SST) was obtained from MODIS (NOAA 2022). Differences in environmental conditions (e.g. SST, Chl a and Kd490) between the sites were analysed using monthly mean values.

DNA amplification and sequencing

The Symbiodiniaceae ITS2 region was amplified with the specific primer pair ITSintfor2 (5′- TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGGAATTGCAGAACTCCGTG-3′) and ITS2-reverse (5′TCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGGGATCCA TATGCTTAAGTTCAGCGGGT-3′) (LaJeunesse 2002) that were used with adapters (underlined) for sequencing on the Illumina platform with a MiSeq V2 500 cycle kit. PCR amplification of ITS2 rDNA was conducted under the following conditions: 15 min at 95 °C, followed by 35 cycles of 94 °C for 45 s, 55 °C for 45 s, 72 °C for 1 min, and final extension of 72 °C for 10 min. PCR product was purified using Agencourt AMPure XP PCR purification beads, following the manufacture protocol. 16 µl of beads were added to each sample to remove 150 bp < according to the calculation of: sample volume X 0.8 = beads volume. Index PCR was then performed with unique combinations of Nextera XT index primers for each sample. Amplicons were cleaned, pooled, and sequenced at Genomics WA facility in Perth.

Bioinformatics

All filtering steps and statistical analyses were conducted in R software, 4.1.1 (R Core Team 2021). Raw sequence data were clustered into amplicon sequence variants (ASV’s) using the DADA2 pipeline (Callahan et al. 2016). We filtered poor quality samples with reads less than one thousand and any ASV’s with less than ten total reads across all samples for downstream analysis. Following quality control, ten samples and six ASV’s across all sites were removed. All sequences were aligned to a reference file consisting of known symbiont ITS2 haplotypes using BLASTn in order to taxonomically classify our unique ASV’s.

Data and statistical analysis

The spatial genetic structure of Turbinaria reniformis coral host samples was determined using SNP dataset derived from Evans et al., (2020). This dataset was processed with the dartR package (Gruber et al. 2018) and visualised through discriminant analysis of principle components (DAPC) using the adegenet package (Jombart 2008).

Difference in symbiont community composition between the study sites and between site by SST/ Chl a/ Kd490 interaction was analysed using two-way permutational multivariate analysis of variance (PERMANOVA) using the Bray–Curtis dissimilarity measure in the vegan package (Oksanen et al. 2012). Specifically, the pairwise.adonis2 function with 999 permutations was applied to explore pairwise variations across the study sites. To investigate, the effect of the environmental data (monthly mean) in driving the Symbiodiniaceae community composition, we performed a distance-based redundancy analysis (dbRDA) using the function capscale followed by an ANOVA to check the model significance. Kd490 and Chl a presented the same relative contribution (28%), thus, we chose to show Kd490 in the plot. Different statistical models (NMDS and PCoA) were employed before ultimately choosing the presented model. All tests yielded similar results, supporting the presented conclusions.

Shannon’s diversity index of the Symbiodiniaceae community was calculated for each site and differences along the latitudinal gradient were tested by a Kruskal-Willis test followed by a pairwise Wilcoxon test with Bonferroni correction method.

To investigate whether geographical distance between sites explained the variability in symbiont communities, we utilized a multiple matrix regression model with randomization (Wang 2013). This analysis employed the fossil package (Vavrek 2011) and included calculations of distance using earth.dist function within the ecodist package (Goslee and Urban 2007). To assess the relationship between coral host and its symbiont community, the mantel function was employed with Pearson’s correlation method and 999 permutations using the vegan package (Oksanen et al. 2012). Lastly, one-way ANOVA was performed followed by a TukeyHSD post-hoc tests to identify if there were significant differences in mean monthly environmental conditions (Kd490, Chl a and SST) between the six sites. All environmental data was Log10 transformed to meet statistical test assumptions.

Results

Environmental data

We identified strong environmental differences among our sites (Table S1). The monthly mean Kd490 data exhibit significant differences among all sites along the latitudinal gradient, except for MI-DA (ANOVA, F = 158.66, p < 0.0001). The northernmost site, BB, recorded the highest mean at 0.23 ± 0.04, while the southernmost site, Perth, recorded the lowest mean value at 0.07 ± 0.01. EG’s values closely resemble those of BB at 0.21 ± 0.04, while the remaining sites display Kd490 values ranging from 0.14 to 0.16 (Fig S1 and Table S2). Likewise, monthly mean Chl a levels exhibit significant variations among all sites, excluding MI-DA (ANOVA, F = 235.23, p < 0.0001). Highest Chl a values were observed at BB (3.26 ± 0.78 mg m3), while Perth had the lowest Chl a at 0.62 ± 0.23 (mg m3). A distinct temperature gradient was also observed across the sites (ANOVA, F = 83.4, p < 0.0001), with Perth, the coolest high-latitude site, displaying a monthly mean SST of 20.77 ± 1.71 °C, and BB, the warmest low-latitude site, recording a monthly mean SST of 27.05 ± 3.73 °C (Fig S1 and Table S1, S2).

ITS2 metabarcoding

Amplicon sequencing of the ITS region returned a total of 6,332,251 reads, with 115,131 mean reads per sample (n = 55; Table 1). These sequences were clustered into 19 different amplicon sequence variants (e.g. lineages), which were further classified into five different Symbiodiniaceae genera (13- Cladocopium, 1- Fugacium, 3- Symbiodinium, 1- Durusdinium, 1- Breviolum; Fig. 2).

Table 1 Summary of ITS2 based ASV data at each site. n = number of samples per site, QC output = sum number of sequences post quality control, mean ASV reads per sample, ASV’s = total number of ASV’s and Shannon diversity index (mean ± sd)
Full size table
Fig. 2
figure 2

Symbiodiniaceae community proportional composition (%), based on ASV read count in each colony (bar), across all sites along the latitudinal gradient. From north (top) to south (bottom)– Balla Balla (BB), Dampier Archipelago (DA), Montebello Islands (MI), Exmouth Gulf (EG), Shark Bay (SB) and Perth (Pe). ASV’s represent Symbiodiniaceae of the genera Symbiodinium (A), Breviolum (B), Cladocopium (C), Durusdinium (D) and Fugacium (F)

Full size image

Symbiont community composition

Although the Symbiodiniaceae genus Cladocopium dominated all colonies across all examined sites (Fig. 2), we found strong differences in Cladocopium lineages among our tropical and higher-latitude sites. At the four Pilbara sites coral colonies had 7–12 ASV’s per sample which were primarily associated with the C33.1a lineage (42.3% at BB, 31% at DA, 25.3% at MI and 20.7% at EG;Table 1; Fig. 2). In SB, the C33.1e lineage dominated, comprising 86.8% of the Symbiodiniaceae type, while Perth displayed a more variable symbiont community with the c.b (26.8%), and C33.1c (12.1%) lineages being the most abundant among samples.

Although all samples were dominated by Cladocopium, other genera within the Symbiodiniaceae family, such as Durusdinium, Symbiodinium, Breviolum and Fugacium, were detected at background levels. In Perth, one sample was associated with clade F (1.46%), and one with clade B18 (64.7%). Three samples from DA were associated with clade D1 (0.002–12%), while the three different sub clade A were found in three different samples from EG (A.a; 0.46%), EG (A.b; 0.01%) and SB (A.c; 0.004%) (Fig. 2).

PERMANOVA analysis revealed significant effect of site (r2 = 0.28, F = 3.89, p = 0.001) on the symbiont community assemblages (Fig. 3, Table S4). SB and Perth were found to have significantly distinct symbiodiniceae communities compared to all other sites, and to each other (r2 = 0.31, F = 5.59, p = 0.004). Additionally,e nvironmental factors showed significant correlation with the variations in the symbiont community composition, as indicated by the dbRDA model: SST (r2 = 0.19, F = 3.29, p = 0.001), Kd490 (r2 = 0.22, F = 3.82, p = 0.001) and Chl a (r2 = 0.22, F = 3.83, p = 0.001). SST explained 43% of the community variation followed by Kd490 (28.6%) and Chl a (28.2%) (Table S3).

Fig. 3
figure 3

Distance based redundancy analysis (dbRDA) for environmental drivers effect on Symbiodiniaceae communities in T. reniformis. Site symbols are colour coded by symbiont community profile. Each point represents an individual colony. The PERMAMOVA result is presented in the top right corner. Kd490 and Chl a presented the same relative contribution, we chose to show Kd490

Full size image

Symbiont community diversity

The Shannon diversity index (Fig S2), calculated for Syimbiodiniceae community profiles in each sample, was significantly different between the sites (Kruskal–Wallis, Chi sq = 33.02, df = 5, p < 0.0001), and in general showed a pattern of reduced diversity in higher-latitude sites (SB and PE). Shark Bay exhibited the lowest community diversity, with 6 ± 2.23 ASV’s per sample (Fig. 4; Table 1). Perth also displayed lower levels of diversity than its tropical counterparts, but with more variable symbiont community profiles than the Shark Bay region (Fig. 4; Table 1).

Fig. 4
figure 4

Shannon Diversity Index of the symbiont community at each site

Full size image

Geographical distance was not found to have a significant effect on the symbiont community composition between the sampled sites (r2 = 0.02, p = 0.679). Furthermore, the host genetic structure and symbiont community composition revealed no significant correlation (R2 = 0.03, p = 0.194).

Discussion

In this study, we investigated the composition of Symbiodiniaceae communities in Turbinaria reniformis along the Western Australian coast, spanning approximately 1,200 km from tropical low-latitude reefs to temperate high-latitude reefs. Our findings reveal that all T. reniformis colonies along the Western Australian coast associate predominantly with Cladocopium. However, within this genus, we found spatial variation of the lineages communities’ composition, influenced primarily by SST, and to a lesser extent by Kd490 and Chl a. Surprisingly, host genetic data did not align with symbiont community divergence, suggesting a complex interplay of environmental factors in shaping these associations.

Latitudinal and environmental differentiation in symbiont communities have been well-documented in studies conducted on various coral species in distinct regions, such as the Red Sea (Terraneo et al. 2019), South China Sea (Chen et al. 2019), and Singapore (Guest et al. 2016). These studies reveal a pattern of transitioning from Cladocopium-dominated symbiont communities in cooler, clearer waters to the stress tolerant Durusdinium in warmer and more turbid environments. However, a study by Matias et al., (2023) examining Acropora tenuis across a latitude gradient in the Great Barrier Reef (GBR) found a consistent high dominance of Cladocopium across the gradient. Here, ASV’s associated with Durusdinium, Symbiodinium, Breviolum, or Fugacium were rare and only identified in nine colonies scattered across all sites. This data suggests that these Symbiodiniaceae genera occur at background levels in the coral Turbinaria reniformis. The role of these background symbionts remains unclear, and under our experimental design, their presence cannot be attributed to any specific environmental driver. Therefore, it is plausible that they represent transient rather than stable symbiont partners (Lee et al. 2016).

It is crucial to acknowledge that different taxa may exhibit varying affinities for specific symbiont genus, emphasizing the need for caution when comparing studies involving different host taxa (Smith et al. 2020). In a prior study in WA by Silverstein et al., (2011), which explored symbiont associations in various coral genera, Turbinaria mesenterina from Ningaloo and T. mesenterina and T. reniformis from Dunsborough, south of Perth, were also found to be predominantly associated with Cladocopium. Similarly, Thomas et al., (2014) observed that Cladocopium was the dominant symbiont in Acropora sp. from two distinct reefs in WA. These findings collectively underscore high stability and specificity to Cladocopium across large geographic distances in WA, despite significant variations in SST and Kd490.

The composition of Cladocopium communities exhibited significant variations between sites and in relation to environmental factors along the latitudinal gradient. Specifically, the four Pilbara sites displayed strong similarities in symbiont community assemblages, with a dominant presence of the C33.1a lineage, indicating a shared bioregion. In contrast, Shark Bay colonies exhibited limited community diversity, lacking the presence of C33.1a, while those in Perth displayed more variable symbiont assemblages, including the highest number of ASV’s among the sites.

While, previous literature lacks data on C33.1, higher SST’s in the Pilbara region might explain the predominant presence of C33.1a in the four northern sites, while the influences of Kd490 and Chl a on symbiont community diversity were comparable but less pronounced. The limited variability in the symbiont community of Shark Bay may be attributed to its unique environmental conditions, as it serves as a transition zone, where the Leeuwin Current meets the Ningaloo Counter-Current, creating a connectivity boundary for particle transport, potentially restricting symbiont distribution (Evans et al. 2019; Woo et al. 2006). Further investigations are warranted in this area, particularly focusing on coral host populations and their associated symbiont communities. Similar patterns were observed in the Persian/Arabian Gulf, where extreme environmental factors, such as high SST and salinity, constrained Symbiodiniaceae diversity, leading to the prevalence of Cladocopium symbionts with high heat and salinity tolerance (D’Angelo et al. 2015; Hume et al. 2016; Ziegler et al. 2017). Additionally, high turbidity was suggested to be responsible for the low richness and diversity of Symbiodiniaceae in Hong Kong coral reefs (Ng and Ang 2016).

Differences between various Cladacopium lineages and their effect on the coral host are still poorly understood. Moreover, coral species with high symbiont specificity may have limited capacity to switch or shuffle symbionts in response to environmental changes (Johnston et al. 2022). Interestingly, although differences were identified in the lineage communities between the regions, this was not attributed to geographical distance or related to the host genetic differentiation. This pattern may be attributed to the reproductive strategy of Turbinaria reniformis, a broadcast spawning coral that obtains its symbionts horizontally (i.e., acquired from the environment) (Baird et al. 2009). Another possibility is that Cladocopium subclade community divergence among reefs might be attributed not only to dispersal limitation but also to environmental specialization and local adaptation of the symbionts (Davies et al. 2020). Despite a lack of relationship between symbiont and host structure, similar overall trends in both the host and symbiont datasets were evident and showed that Perth and SB were highly diverged from their tropical counterparts to the north. The strong structure in the host within the Pilbara, and the lack of differences in symbiont communities among these sites, was likely the driver of the non-significant relationship between the host genetic structure and the symbiont communities. Nevertheless, the pronounced genetic divergence of the host at higher-latitude sites aligns consistently with the patterns of symbiont community composition identified in this study. This observation suggests that T.reniformis coral host have some flexibility in their symbiotic associations with Cladocopium across their range and within their specific environments (Berkelmans & Van Oppen 2006; Davies et al. 2020).

The genus Cladocopium is highly diversified, a fact that has been associated with functional variation in symbiont thermal tolerance across reefs as well as with functional differences between reef zones, providing support for the potential for reef-specific symbiont communities (Barfield et al. 2018; Davies et al. 2020; Howells et al. 2011). We hypothesise that Turbinaria sp. widespread nature is thus attributed to its strong association with Cladocopium. This insight is crucial for planning conservation strategies in extreme or marginal reefs. We also highlight that coral reefs of WA offer an excellent study system for investigating fine-scale local adaptation potential of Cladocopium and its coral hosts.

Further in-depth investigations are needed for other generalist and resilient species, to substantiate such assumptions. This becomes even more pertinent considering turbid reefs may serve as refugia under climate change (Cacciapaglia & van Woesik 2016; Morgan et al. 2017). As we progress in researching marginal reefs and their potential role as refugia, integrative molecular approaches that consider both Symbiodiniaceae and host population structure will be imperative. These approaches will provide a more holistic understanding of how different life history strategies, including reproductive and symbiont acquisition modes, influence the connectivity levels between neighbouring coral habitats.