Introduction

Gracilaria production has become a significant component of the global aquaculture industry. With its versatile applications in food, pharmaceuticals, cosmetics, and biofuel, Gracilaria cultivation has expanded worldwide (Ferdouse et al. 2018). Several countries have emerged as key players in Gracilaria production, including China, Indonesia, the Philippines, Vietnam, and Chile (Mantri et al. 2023). As global demand for seaweed products continues to rise, the production of Gracilaria is poised to play an increasingly vital role in meeting these growing needs while promoting economic development and environmental sustainability. Concomitantly, Gracilaria also represent a valuable resource for local communities, as its farming is carried out mainly by local fishermen, contributing directly or indirectly to meet at least five independent United Nations SDGs goals (goals 1, 5, 8, 12 and 14; see https://sdgs.un.org/goals).

Gracilaria chilensis (a.k.a. pelillo) is the most produced seaweed in Chile via aquaculture (Buschmann et al. 2017). Chilean farmers have adopted both traditional direct “off-bottom” planting (Buschmann et al. 1995) and innovative farming methods, utilizing coastal ponds (Santelices and Doty 1989), long-lines (Westermeier et al. 1993), indoor (Caroca-Valencia et al. 2023) and integrated multi-trophic systems (Abreu et al. 2009) to grow this seaweed. Historically, pelillo farming in Chile has faced recurrent pest outbreaks which stagnate the expansion of the Chilean industry (Mendez et al. 2024). Pests such as epiphytic organisms, fouling species, grazers, and pathogens can compete with Gracilaria for resources, inhibit its photosynthetic efficiency, and cause physical damage to the seaweed fronds (Buschmann et al. 1997; Leal et al. 2020). As a result, the affected G. chilensis may exhibit stunted growth, decreased reproductive output, and increased susceptibility to diseases (Fletcher 1995). Moreover, pest and disease outbreaks can escalate rapidly, spreading between farms or wild stocks, exacerbating the ecological impact and posing a significant biosecurity risk to the surrounding marine environment (Murúa et al. 2023).

A poorly studied aspect on seaweed farming is the interaction of cultivated stocks with local ecosystems, for example pest ecology. In terms of Gracilaria pests, there is some knowledge of algal pests being controlled by grazers such as polychaetes controlled by foraging seabirds (Martínez-Curci et al. 2023). Nevertheless pathogen-mediated biocontrol is much as unclear as the general knowledge of algal disease in Latin America (Murúa et al. 2021). In recent observations within a G. chilensis suspended system, we identify a fungal-like outbreak that unexpectedly acts as a natural biocontrol agent, effectively suppressing a prevalent seaweed pest population of Porphyra. We identified the oomycete species responsible for such pest control, their interactions with the targeted pest, and the potential implications for sustainable Gracilaria farming. By exploring the ecological factors contributing to the fungal outbreak and its impact on pest populations, we hope to gain valuable insights into the natural biocontrol mechanisms in seaweed farming systems. This research not only highlights the potential of utilizing natural agents for pest management but also offers prospects for developing eco-friendly strategies to enhance the resilience and productivity of Gracilaria cultivation.

Materials and methods

Sampling

A suspended Gracilaria chilensis set-up was installed in Cariquilda estuary (41.624°S; 73.591°W: Fig. 1A) in spring 2022 (28–09-2022), in order to evaluate epiphyte succession of pests on a biweekly basis. Five 5 m long long-lines (n = 5) seeded at ca.120 g m−1 were installed superficially (1 m depth). From every long-line, 30 cm were harvested biweekly for Gracilaria biomass and pest load estimations. In particular, we focused on epiphytic Porphyra (Fig. 1B), which was estimated by i) its coverage on 50 randomly harvested Gracilaria filaments measured linearly on a millimeter paper and ii) the epiphyte weight in terms of 100 g of its basiphyte. As a matter of fact, both proxies are correlated (R2c = 0.999; p = 0.003; Fig. 2A), and coverage will be used for Porphyra abundance hereafter. Additionally, fifteen Porphyra individuals per sample were dissected for the estimation of the prevalence of the fungal-like disease.

Fig. 1
figure 1

Identification of the red rot disease agent Pythium porphyrae on Gracilaria-epiphytic Porphyra sp. A: Gracilaria chilensis farmed on a suspended long-line in the Cariquilda estuary. B: macroscopic view of Gracilaria filaments colonized by Porphyra recruits (arrowheads). C: Porphyra sp. symptomatic for a Pythiosis (a.k.a. red rot disease in case of Bangiales, arrows). D: upon magnification, necrotic areas show the infection by filamentous fungal like structures, which introduce within Porphyra tissues and cells. Inset: after calcofluor white staining (CFW), pathogen cells can be easily spotted inside the host tissue under UV light. AUTO: Chlorophyll autofluorescence. E: in older cultures, P. porphyrae overgrew Porphyra blades (arrows) and cover most of the tissues. Inset: Pythium sporangia (arrowheads) in heavily infected tissues. F: A Pythium porphyrae isolate growing in CMA culture media

Full size image
Fig. 2
figure 2

Variations of Porphyra epiphytic load and Pythium prevalence during a Gracilaria chilensis suspended set-up. A: both Porphyra coverage and weight are strongly correlated, after a Generalised linear mixed model including the long-line sampled and the week as random effects. B: Porphyra abundance variation in 8-week Gracilaria farming. C: Pythium prevalence variation (% diseased Porphyra specimens) over the same 8-week course. Both weeks and and Porphyra coverage were significant for the GLMM model fit

Full size image

Statistical analyses were carried out using R (R Core Team 2018). To demonstrate a potential link between coverage and weight of Porphyra recruits, data were fit to a generalised linear mixed model (GLMM) with Poisson error structure using the {glmer} function of the “lme4” package, where long-line measured and time (n° week) were set-up as random effects (Bates et al. 2015). Likewise, to assess the association Pythium prevalence and Porphyra coverage or week, another GLMM was performed with Poisson error structure and long-line as random effect model. Pairwise comparisons were done applying a Tukey test on the GLMM models using the library “lsmeans” (Lenth 2016). Both conditional (r2c, fixed, and random effects) and marginal (r2m, fixed-factors effects) goodness of the fits were calculated to estimate the impact of our random effects (Johnson 2014).

Oomycete isolation

Porphyra blades symptomatic for the red rot disease were acquired from our Gracilaria farm facility by 10–11-2022. These samples were carefully stored in 50 mL Falcon tubes in a cooled container with ice packs and transported to the FICOPAT lab (UACh) in Puerto Montt, Chile. To isolate the pathogen, we placed infected tissues (ca. 3 mm diameter each) on 1.7% corn meal agar (CMA BBL) plates, prepared in filtered (0.2 µm) and autoclaved seawater from the estuary with a salinity of 25 PSU. Additionally, before pouring the agar, we supplemented with 200 mg mL−1 of penicillin and streptomycin sulphate to prevent bacterial contamination (Atami et al. 2009).

The agar dishes were sealed using Parafilm and then incubated at 12 °C in darkness. They were monitored daily for hyphal growth. After 14 days incubation, single hyphal tips were excised out the agar cultures and transferred to a fresh CMA in Petri dishes, to establish pure cultures. As these visible mycelia grew, we aseptically subcultured single hyphal tips by transferring them to Petri dishes with fresh medium. This process was repeated several times until the cultures were completely free of any visible bacterial or protist contamination.

Microscopy and molecular diagnoses

Microscopical observations of diseased Porphyra and oomycete isolates were performed on fresh material. Samples were mounted using a Calcofluor white (CFW) solutions (0.02 mg mL−1 in seawater; Gachon et al. 2017), observed and imaged on a Zeiss Microscope Axioscope 5 DIC and epifluorescence microscope with a Axiocam 202 Color digital camera, under a Colibri 3 Illumination System.

For molecular diagnoses, samples were DNA-extracted using a commercial kit (GeneJet DNA extraction kit, Thermofisher). PCRs were performed to amplify 18S (primer pairs F139/R1233 complemented with EUK422-445/EUK1422-1440_R and TK18Sfwd/ TK18Srev, annealing = 55 °C), COI (primers TKCOIfwd -TKCOIrev, annealing = 55 °C) and COII (primers COX2-For3 – COX2-Rev3, annealing = 55 °C) regions (Sekimoto et al. 2007, 2008; Wang et al. 2014; Gachon et al. 2017; Klochkova et al. 2017b), using reagent concentrations established in Murúa et al. (submitted), and sequenced by Sanger (Macrogen). Phylogeny reconstructions were inferred by aligning Pythium clade I and II sequences available in NCBI in addition to those newly generated by this study (accessions: PP481400/PP481401 for 18S, PP503415 for COI and PP493240/PP493240 for COII). Alignments were performed using MAFFT (Katoh and Standley 2013) and modelled for phylogeny reconstruction using RaXML version 8.2.11 with 1000 bootstrap replications (Stamatakis 2014) and MrBayes version 3.2.6 using default settings four MCMC chains (Ronquist et al. 2012), implemented in Geneious v11 (Kearse et al. 2012). Analyses were made for each gene separately.

Results

Once the Gracilaria set-up was installed (Fig. 1A), Porphyra recruitment started right away after two weeks and progressively increased to reach a peak after 6 weeks of cultivation (50–72%) (Figs. 1B, 2B). At the 8th week, there was a collapse of epiphytic Porphyra on Gracilaria thalli, with no visible recruits or elongated Porphyra from previous months.

Concomitantly, Porphyra started to show changes in coloration over the course of cultivation. Whilst 2-weeks-old Porphyra looked dark red, 4- and 6-weeks old specimens show irregular pink areas across the thalli, of variable irregular areas (Fig. 1C). Under the microscope, such areas evidenced an intracellular hyphal organism infecting the host cells (Fig. 1D), which contained a cellulosic/quitinated cell walls and caused necrosis in neighbouring host cells. In very infected host tissues (ca. 6 weeks old), such hyphae emerged and formed extra-tissular complexes that included sporangia of 15–20 µm diameter in their tips (Fig. 1E). The prevalence of diseased Porphyra follows the same pattern of rise and fall of its recruitment (Fig. 2C), with symptomatic individuals detected after 4 weeks (up to 12%), reaching a peak of infection (60–100%) at 6 weeks, and no records at 8 weeks.

The symbionts were culturable in solid medium (CMA medium; Fig. 1F). We obtained two fungal-like strains from infected Porphyra (Pyt1 Cariq and Pyt2 Cariq), which were subjected to molecular phylogeny. Both isolates were identical for 18S sequences and differed in only two bases for COII nucleotide sequences. The phylogenetic analyses, under both RAxML and MrBayes, showed that Chilean samples formed a well-supported clade within Pythium clade A, using 18S, COI and COII (Figs. 3, 4 and 5). Within this clade, our isolates share monophyly with Pythium chondricola, P. porphyrae, and to a lesser extent P. adhaerens. Members of P. porphyrae includes specimens from Japan, China, Korea and New Zealand, suggesting a cosmopolitan distribution of the species.

Fig. 3
figure 3

Bayesian phylogenetic tree reconstruction of Chilean Pythium porphyrae (in bold pink within the tree) and available Pythium SSU rDNA sequences. The tree contains a total of 23 sequences and 1834 nucleotide positions. Support values given are posterior probabilities (MrBayes)/bootstrap support (RAxML). MrBayes settings: chain length 1.000.000, subsample frequency 1.000, burn in of 10%. The scale bar indicates the number of substitutions per site

Full size image
Fig. 4
figure 4

Bayesian phylogenetic tree reconstruction of Chilean Pythium porphyrae (in bold pink within the tree) and available Pythium COI mtDNA sequences. The tree contains a total of 26 sequences and 682 nucleotide positions. Support values given are posterior probabilities (MrBayes)/bootstrap support (RAxML). MrBayes settings: chain length 1.000.000, subsample frequency 1.000, burn in of 10%. The scale bar indicates the number of substitutions per site

Full size image
Fig. 5
figure 5

Bayesian phylogenetic tree reconstruction of Chilean Pythium porphyrae (in bold pink within the tree) and available Pythium COII mtDNA sequences. The tree contains a total of 22 sequences and 615 nucleotide positions. Support values given are posterior probabilities (MrBayes)/bootstrap support (RAxML). MrBayes settings: chain length 1.000.000, subsample frequency 1.000, burn in of 10%. The scale bar indicates the number of substitutions per site

Full size image

Discussion

Pest outbreaks are keystone factors regulating profitability and sustainability in seaweed aquaculture (Brakel et al. 2021; Murúa et al. 2023). In G. chilensis farms, pests such as worms (Polychaeta), ceramialean blooms or green tides have historically had significant consequences on both the cultivated seaweed, the overall aquaculture operations, and the social acceptability of surrounding settlements (Mantri et al. 2023). For instance, these outbreaks end up with Gracilaria reduced biomass, mixed with contaminants (e.g., pests), stunted morphologies and sometimes diminished biochemical quality; thus, economic detriment for farmers (Fletcher 1995). Furthermore, major decomposing pest drifts may lead to human health problems. Such consequences underscore the importance of effective pest identification and management strategies, including early detection, proper sanitation measures, and the implementation of biosecurity protocols (Mendez et al. 2024). By preventing and mitigating the impact of pest outbreaks, farmers can safeguard the health and productivity of seaweed farms, ensuring the sustainability and success of valuable seaweed aquacultural species (Cottier-Cook et al. 2021, 2022).

In Chile, G. chilensis has been extensively farmed in systems with direct attachment to substrata (Buschmann et al. 1995). Nevertheless, suspended lines have drawn some attention, as the cultivable areas can be expanded from tidal-affected regions towards offshore. Suspended systems had suggested a lower epiphytic load than direct attachment (Westermeier et al. 1993), in contrast to abundant filamentous red and brown seaweeds that are very common in traditional systems (Leonardi et al. 2006). Nonetheless, a different bathymetrical level potentially incorporates new ecological interactions. For example, Porphyra species had not been recognized in off-bottom systems as a significant epiphytic pest affecting Gracilaria farms (Candia et al. 2006). Nevertheless, in our study Porphyra was one of the most abundant epiphytes in suspended systems during 2022 and 2023 (unpublished). It is unknown how Gracilaria would be affected in terms of growth, survival or biochemical quality (e.g. agar composition) in case Porphyra would keep growing, although due to its superficial attachment (type 1–2 after Leonardi et al. 2006), it is presumed a significant competitive interaction only when blades are considerably larger. On the other hand, G. chilensis was not observed affected by P. pophyrae. Despite its initial abundance, Porphyra co-existence in Gracilaria farms is rather ephemeral, suggesting an outward disruption source during Porphyra recruitment and settlement during the initial weeks.

In this system we report a third party: a fungal-like organism inhabiting Porphyra, whose prevalence was directly linked with the demise of Porphyra recruits. Identified as Pythium porphyrae after marker-assisted taxonomy, the facultative parasitic oomycete caused pink lesions and massive mortality in Porphyra blades, resembling red rot disease symptoms (Klochkova et al. 2017a). Our molecular data suggests that Chilean Pythium is very similar genetically to those found on Pyropia plicata in New Zealand (Diehl et al. 2017) and Porphyra and Pyropia spp. in China, Korea and Japan (Kawamura et al. 2005; Kim et al. 2014; Qiu et al. 2019). Our phylogenetic analyses also suggest the conspecificity between P. porphyrae and P. chondricola (Diehl et al. 2017), with neither morphological nor genetic dissimilarity except for COII, which are rather different between them (Robideau et al. 2011; Lee and Lee 2022). Both species have been isolated primarily from algae (Lee et al. 2015), and can 100% identical using other sequences such as ITS1 (Lévesque and De Cock 2004). More powerful techniques (e.g., phylogenomic, phenomics) will be requested in order to pinpoint the specific genetic and phenotypic differences between both traits and to know whether they are currently under a speciation process.

The three-party interaction described here raises important questions about the utilization of microbes in seaweed pest biocontrol. Because of the fragility and dynamisms of the aquatic environment, it is highly non recommendable to use chemical methods for pest control as in terrestrial agriculture (Pawan Kumar et al. 2023). Managing pest suppression by the addition of specific microorganisms or complete microbial communities resulted as an eco-friendly alternative of controlling unwanted organisms (O’Hanlon et al. 2012). In forestry, it is increasingly common the use of native entomopathogenic fungi, bacteria and viruses for the natural biocontrol of insect pests (Dara et al. 2019). In aquatic systems, seaweed fitness has been improved by other accompanying organisms. For example, some mesograzers increased Gracilaria foliifera fitness by eating epiphytes that compete with their host (Brawley and Adey 1981). Similarly, amphipods reduced epiphytic biomass from the seaweed Sargassum filipendula (Duffy 1990). As bioengineers. Gracilaria also increase the invertebrate biodiversity in the faring area, increasing worm pests that are controlled by migratory birds (Martínez‐Curci et al. 2023). In our study we describe a microorganism regulating a macroalgal pest management of a commercially important seaweed. In spite of its potential use for biocontrol in Gracilaria aquaculture, its use must be carefully assessed in terms of suitability and environmental impacts.

In contrast to long-lasting pests, Porphyra was quickly controlled by the naturally-occurring P. porphyrae, counteracting Porphyra prevalence to non-detectable levels and indirectly aiding farmed Gracilaria. Nevertheless, P. porphyrae has been described as a significant biosecurity threat on Pyropia and Porphyra farms in Asia (Badis et al. 2020). For the first time, we report this parasite in Chilean waters, proposing a widespread distribution alongside subantarctic ecosystems. These records add up to latest records of another pathogenic oomycete of Bangiales, such as Olpidiopsis (sensu lato) porphyrae in the Southeastern Pacific (Murúa et al. submitted). Both parasites affect farmed and cultivated specimens, whose impact in natural populations or incipient Pyropia aquaculture is yet to be assessed.