Introduction

Obligate root parasitic plants of the Orobanchaceae family (broomrapes) have, through evolution, lost their ability to perform photosynthesis and now rely on parasitizing the roots of other flowering plants to survive1. There are approximately 4500 known species of parasitic plants; of these, nearly 60% are root parasites, while the remainder parasitize the aerial parts of plants. Parasitic plants are known for producing numerous tiny seeds, some of the smallest in the plant kingdom2,3. Orobanchaceae is the largest parasitic plant family, with 102 genera and over 2100 species4. Several broomrape species parasitize significant crop plants, leading to substantial yield losses, particularly in North and East Africa, Southern and Eastern Europe, and the Middle East. Phelipanche ramosa (L.) Pomel is one of the most economically significant broomrape weeds, and a widespread species, attacking various crops in the Solanaceae family, especially tomatoes and tobacco (Fig. 1A–C), as well as crops in the Brassicaceae, Cannabaceae, Fabaceae, Apiaceae, and Asteraceae families. It can reduce crop yields by 2–80%, and in some cases, may cause complete crop failure5. Phelipanche ramosa is a slender, branched holoparasitic plant with a glandular-pubescent stem, usually up to 30 cm tall, and displays a color range from yellowish to brown-violet or bluish hues (Fig. 1C, G)6. Its inflorescence, occupying about half of the stem length, bears bluish to violet flowers, rarely yellow or white, with dark veins (Fig. 1H) and a distinct oval to hemispherical stigma that is white or bluish and approximately 1.5 mm long (Fig. 1I)6,7,8,9. The species parasitizes a broad spectrum of hosts and is distributed throughout Europe, North Africa, and the Middle East, extending to Asia and naturalized in the Americas10. It commonly occurs in lowland agricultural and ruderal habitats with nutrient-rich sandy or loamy soils7. Nicotiana tabacum L. (tobacco), a Solanaceae crop cultivated on over 4.2 million hectares worldwide, is economically important for cigarette, cigar, and chewing tobacco production and is increasingly valued as a biofuel and bioethanol source11,12. China is the main producer, followed by India, Brazil, Zimbabwe, and the USA. In addition to its industrial relevance, N. tabacum serves as a model organism in biotechnology, genetics, and plant pathology research11,12. The annual plant grows 1–3 m tall and is covered with sticky trichomes. Its red, pink, or white tubular flowers form paniculate inflorescences (Fig. 1D, E) and contain a single pistil with a green stigma (Fig. 1F) and five stamens13,14.

Fig. 1
figure 1

Studied holoparasitic Phelipanche ramosa and host-tobacco Nicotiana tabacum: (A) tobacco plantation in Brzeziny (Poland); (B) numerous parasites in tobacco cultivation; (C) P. ramosa attached to the roots of the host-tobacco; (D–F) inflorescence, flower and stigma of tobacco; (G–I)—inflorescence, flower and stigma of P. ramosa. Scale bars: 1000 µm (F), 500 µm (I).

Full size image

Besides causing direct damage, parasitic plants can indirectly affect their hosts by transmitting a wide range of microorganisms, including bacteria, fungi, viruses, and phytoplasmas2. These plants develop specialized feeding structures known as haustoria, which penetrate the host vascular tissue and create a physiological bridge, allowing bidirectional movement of water, nutrients, carbohydrates, small molecules such as RNA and proteins, and microbes2,15,16,17,18. For instance, the bacterial communities of the root holoparasite Orobanche hederae Duby suggest that while the microbiome of parasitic plants is largely derived from their hosts, it remains distinct in structure and function, reflecting the unique ecology and heterotrophic lifestyle of holoparasites17. Among the microorganisms transmitted via parasitic plants are not only pathogens but also beneficial microbes with plant growth-promoting properties19,20,21,24,25. This highlights that the microbial flow between parasites and hosts plays a more complex ecological role than previously assumed. Transmission routes include the haustorium as well as arthropod-mediated transfer to above-ground tissues, including flowers and pistil stigmas19,20,21, and seed-borne dissemination22,23,24,25. Because Orobanchaceae can produce up to 600,000 seeds26, these seeds may act as long-term reservoirs of diverse microbial communities22,23,24,25,27. To mitigate the impact of parasitic plants on crops, biocontrol strategies are being explored that take advantage of the natural movement of microbes between parasites and their hosts. Several studies have demonstrated that antagonistic microorganisms can suppress parasitic plants and therefore serve as potential biocontrol agents. For example, Fusarium toxins can inhibit P. ramosa germination28,29, while metabolites from Myrothecium verrucaria and Fusarium compactum reduce P. ramosa seed germination30. Certain strains of Fusarium are highly pathogenic to parasitic plants such as P. aegyptiaca, P. ramosa, and O. cumana31, and endophytic bacteria such as Pseudomonas sp. PhelS10, isolated from tomato, can decrease P. aegyptiaca parasitism, suggesting potential biocontrol applications16. Therefore, identifying groups of potentially antagonistic microorganisms, which is one of the goals of this paper, will support future targeted research on the biocontrol of parasitic plants and help narrow the search for effective candidates in the future, both in our studies and in the work of other research groups.

The stigma, a key part of the pistil, plays a central role in pollination and fertilization by serving as the initial site for pollen germination and pollen tube growth32. Its surface chemistry, rich in enzymes, sugars, amino acids, glycoproteins, and phenolics33,34, supports compatible pollen while influencing microbial colonization through both attraction and selective inhibition. The stigma’s nutrient-rich and humid environment creates a specific niche that favours adapted microorganisms. Epiphytic microbes inhabit its surface, whereas endophytes colonize internal tissues. Compared with the phyllosphere, the stigma offers greater nutrient availability and moisture but stronger selective pressures due to its reproductive function. Flower parts, including stigmas, are valuable nutrient sources for microorganisms, as reflected by their high microbial abundance and diversity19,20,21,35,36,37,38. Microbes such as bacteria and yeasts metabolize floral nutrients, modifying nectar composition39, while some, for example Acinetobacter spp., can degrade pollen grains to access internal nutrients40.

Holoparasitic plants represent an extreme form of parasitism, completely dependent on their hosts for nutrients and resources. Despite their ecological and economic significance, the microbiomes of holoparasitic plants remain largely unexplored, and even less is known about how these microbial communities interact with those of their host plants. Understanding these relationships is essential to reveal how parasitism shapes plant-associated microbiota and whether holoparasites harbor distinct microbial assemblages that may influence their physiology or host colonization success. Therefore, the aim of this study was to determine the composition of the microorganisms (bacteria and fungi) of the flower stigma and to assess the biodiversity and abundance of their microbial communities, including those on mature pistil stigmas from the parasite plant P. ramosa and its host N. tabacum, both flowering at the same time. Specifically, we asked: How does the stigma microbiome of a holoparasite differ from that of its autotrophic host, and what functional traits distinguish these communities? We investigated the microbiological characteristics of the stigma by identifying both differences and similarities between species using molecular techniques, specifically Next Generation Sequencing (NGS). Furthermore, we analyzed the relationships between microbiome composition and its function (biochemical and physiological traits). Finally, based on the identified bacteria and fungi and their function, we assessed microbes for their potential pathogenicity, and plant growth-promoting potential for the studied species. To the best of our knowledge, this is the first attempt to characterize microbial communities of the stigmas of parasitic plant P. ramosa and its host.

Materials and methods

Plant material and collecting samples

Experimental research and field studies on plants, including the collection of plant material, complied with relevant institutional, national, and international guidelines and legislation, including the Act on Nature Conservation (Dz.U. 2004 nr 92 poz. 880, with later amendments) in Poland, the Regulation of the Minister of the Environment of 9 October 2014 on the protection of plant species (Dz.U. 2014 poz. 1409) in Poland, regulations of the Regional Directors for Environmental Protection in Poland, as well as we are complying with the IUCN41 and the CITES42. Necessary permits were obtained. Plant material sampling was carried out under permission no. WPN.6400.1.2021.KWAW and WPN.I.6400.3.5.2022.AD. from the Regional Directors for Environmental Protection in Poland and was collected in September 2021. Phelipanche ramosa is partially protected in Poland, and only a select number of individuals have been authorized to conduct research for this project. In our study, we selected a population from Brzeziny (Fig. 1), GPS: 50° 30.0' N; 22° 59.0' E (the southeastern part of Poland), which is among the largest in terms of numbers in Poland. The voucher specimen was stored in the Herbarium of Jan Kochanowski University in Kielce (KTC), Poland43. The collection is not numbered, but the herbarium sheets are stored in a separate section under the name “Parasitic plants”, and the species are sorted alphabetically, in boxes labelled with the species name. Renata Piwowarczyk and Karolina Wiśniewska identified the plant specimens. The plant names were updated according to The International Plant Names Index (IPNI)44.

The flowers of P. ramosa and N. tabacum were carefully placed inside sterile plastic tubes. Stigmas for both plants were collected at the same time and developmental stage, mature in open flowers (MA), and stored at 4 ± 0.5 °C until analysis, which occurred within 24 h. Each sample weighed approximately 1 g and included 350 stigmas of P. ramosa and 200 stigmas of N. tabacum from fully opened flowers (mature stigmas, MA). The total number of flowers collected corresponded directly to the number of stigmas analyzed, as each flower contained a single stigma. The same number of flowers and stigmas was collected for each biological replicate, resulting in three replicates per plant species used for microbiome analysis. The stigmas were dissected using sterile tools, immediately placed into plastic tubes, and stored at −80 ± 0.5 °C for culture-independent microbial analysis via Next Generation Sequencing (NGS). The pistils were removed as close to the stigma’s base as possible. For NGS analysis to assess microbiome composition, a total of 11 samples were analyzed: six samples containing P. ramosa mature stigmas (three for bacterial and three for fungal analysis) and five samples containing N. tabacum mature stigmas (three for bacterial analysis and two for fungal analysis, marked as NT). All available samples were profiled for both 16S (bacterial) and ITS (fungal) analysis; however, one N. tabacum sample intended for fungal analysis could not be identified, resulting in fewer fungal replicates for this species. The stigmas were collected from both the parasite and the host simultaneously, which is notable because these plants often do not bloom at the same time.

Sequencing

The composition of the microbiome of stigmas of P. ramosa and N. tabacum was assessed using high-throughput Next Generation Sequencing (NGS). Molecular analyses were conducted following the methods described by our previous works Ruraż et al.19,20 and Wiśniewska et al.21.

Genomic DNA was extracted from pooled stigma tissue samples (1000 mg) using the Genomic Mini AX Bacteria + kit (A&A Biotechnology, Poland) with an additional mechanical lysis step in a FastPrep-24 instrument (MP Biomedicals, USA) using zirconia beads. DNA was further purified using the Anty-Inhibitor Kit (A&A Biotechnology, Poland) and quantified with a Qubit 4 Fluorometer. The final eluate of DNA had a minimum concentration of 0.1 ng/μL and a minimum volume of 20 μL per sample. The presence of bacterial and fungal DNA was confirmed by real-time PCR targeting the 16S rRNA gene and ITS regions using SYBR Green dye with universal primers.

Before preparing the V3–V4 and ITS amplicon libraries, DNA eluates were assessed for quantity and quality. The 16S V3-V4 region for bacteria and ITS region for fungi were amplified following Illumina’s 16S Metagenomic Sequencing Library Preparation Guidelines (Part #15,044,223 Rev. B) using a two-step PCR protocol with the Herculase II Fusion DNA Polymerase and Nextera XT Index Kit V2. Library quality was evaluated following the Illumina qPCR Quantification Protocol Guide. Amplicons were ligated with Illumina adapters and indexed according to manufacturer protocols. Sequencing was performed by Macrogen (The Netherlands) on an Illumina MiSeq platform using paired-end 2 × 300 bp reads (v3 chemistry, 600 cycles), generating up to ~ 100,000 reads per sample.

DNA extraction and PCR amplification were performed by A&A Biotechnology (Poland), while Macrogen (Netherlands) handled the NGS library preparation and executed Illumina and Sanger sequencing for the 16S rRNA and ITS products. Microbial populations in the samples were investigated by sequencing the V3-V4 region of the 16S rRNA gene and the ITS region. Gene segments were amplified using PCR primers designed for the Illumina methodology. Specifically, primers ITS3F (GCA TCG ATG AAG AAC GCA GC) and ITS4R (TCC TCC GCT TAT TGA TAT GC) were used for the fungal ITS library, and 341F (CCT ACG GGNGGC WGC AG) and 805R (GAC TAC HVGGG TAT CTA ATC C) for the bacterial 16S rRNA libraries. Library quality was checked using the Illumina qPCR Quantification Protocol Guide and the 2200 TapeStation System (Agilent, United States). Sequencing was performed on an Illumina MiSeq PE300 platform.

Statistical calculation and data analysis

Quality control (QC) was applied to each file, including quality filtering (removing sequences with ≥ 5 ambiguous base pairs) and length filtering (removing sequences with a length ≥ 2 standard deviations from the mean). Sequences with fewer than three reads (singletons) or an abundance less than 0.0005% were removed after generating the OTU table. Sequencing data were processed using QIIME 2. Illumina metagenomic datasets were submitted to NCBI’s Sequence Read Archive (SRA) under project number PRJNA1092550 (accession from SAMN44245214 to SAMN44245224). Results were obtained using QIIME 245 and a pipeline based on the SILVA (version 138.2)46 or UNITE (version 04.02.2020)47 databases.

The OTU compositions of bacteria and fungi were analyzed. Dominance classes, representing the proportion of the most abundant taxa in each sample, were determined based on previous research19,20,21. Only OTUs present at ≥ 1% in at least one sample were considered; rare OTUs, singletons, and doubletons (< 0.05%) were excluded from analyses to reduce noise. In Tables 1 and 2, taxa were visually categorized according to dominance classes based on their relative abundance. In Table 1, purple indicates eudominant taxa (35.01%), dark blue dominant (10.01–35%), blue subdominant (5.01–10%), light blue rare (2.01– 5%), grey occasional (1.01–2%) and white casual (1%). In Table 2, red corresponds to eudominant taxa (35.01%), orange to dominant (10.01–35%), yellow to subdominant (5.01–10%), light orange to rare (2.01–5%), grey to occasional (1.01–2%) and white to casual taxa occurring below 1%. The richness and diversity of microorganisms in the samples were assessed using the Shannon diversity index, Pielou’s evenness index, and Simpson’s dominance index. Differences in diversity and evenness indices (Shannon H′, Simpson λ, and Pielou J′) between P. ramosa and N. tabacum stigmas were tested using the non-parametric Mann–Whitney U test.

Table 1 Bacterial community composition and diversity indices identified in Phelipanche ramosa (PR) and Nicotiana tabacum (NT) in mature stigmas (MA) from opened flowers.
Full size table
Table 2 Fungal community composition and diversity indices identified in Phelipanche ramosa (PR) and Nicotiana tabacum (NT) in mature stigmas (MA) from opened flowers.
Full size table

Agglomerative Hierarchical Clustering (AHC) was performed on the full OTU table using Bray–Curtis distance as the dissimilarity measure, and dendrograms were constructed with Ward’s linkage method. Microbial community composition dissimilarity between the two groups was assessed with nonparametric permutational multivariate ANOVA (PERMANOVA) using the adonis function. The Statistical Analysis of Metagenomic Profiles (STAMP)48 was used to identify differentially abundant microbial taxa between the stigma groups of N. tabacum (group 1) and the parasitic plant P. ramosa (group 2), using Welch’s t test with p values corrected by the FDR method.

The MACADAM database49 was utilized to examine the predicted functions of the microbial community, using the MetaCyc subdatabase for Plant Growth Promoting (PGP) properties50 and FAPROTAX for metabolic functions51. Fungal trait analysis was conducted using the FungalTraits database52. The calculations were carried out with XLSTAT v2024.4.1 and XLSTAT-R Vegan53, while the Principal Component Analysis (PCA) was conducted using PAST 4.0354. PCA was conducted on a Pearson correlation matrix, which automatically standardizes data (mean = 0, SD = 1) and reduces multicollinearity. The Kaiser–Meyer–Olkin (KMO) measure was 0.705, indicating moderate to good suitability of the dataset, and almost all variables showed the highest cos2 values on F1 or F2 axes, except for Mucor, where F1 and F3 were similar.

To construct the co-occurrence networks, the data were first processed by generating a Pearson correlation matrix. From this matrix, only pairs showing strong correlations (r = 0.9–1.0) were extracted. The resulting data were then imported into Gephi for network visualization. The linLog layout was applied, and the attraction value was relaxed twofold to optimize network clarity and node separation. Network co-occurrence analysis was performed using the ForceAtlas2 algorithm with LinLog mode55 based on standardized data (n standardization) of Pearson’s correlation matrix in Gephi 0.956.

Partial Least Squares Confirmatory Factor Analysis (PLS-CFA), a variance-based method within the PLS-PM framework, was applied to examine mutual relationships among groups of variables, including bacteriobiome, mycobiome, diversity indices, and trait categories. PLS-CFA is robust to small sample sizes and does not rely on classical regression fit indices, making it suitable for the limited dataset in this study.

Results

Ecological indices and characteristics of the microbial community

For bacterial communities, the Simpson’s dominance (λ), Shannon diversity index (H′), and Pielou’s evenness index (J′) on the stigmas of P. ramosa and N. tabacum reached similar values (λ: 0.3 vs. 0.3, H′: 2.0 vs. 2.0, J′: 0.2 vs. 0.3), with none of the differences being statistically significant (p = 0.100 for all indices) (Table 1). For fungal communities, P. ramosa samples showed a lower Simpson’s dominance (λ) index (0.1 vs. 0.3, p = 0.200) and a higher Shannon diversity index (H′) (2.8 vs. 2.0, p = 0.045) compared to N. tabacum stigmas. Pielou’s evenness index (J′) was slightly higher for P. ramosa (0.3 vs. 0.2, p = 0.810).

The results of the Bray–Curtis dissimilarity analysis presented in the AHC dendrogram for bacteria showed two distinct groups of pistil stigmas which were identified, corresponding to the stigmas of N. tabacum (NT) and P. ramosa (PR). NT MA2 and NT MA3 are more like each other than NT MA1, while PR MA1 and PR MA2 are more like each other than PR MA3 (Fig. 2). Analysis of the rarefaction curves shows that bacteria inhabiting the stigmas of P. ramosa in all analyzed samples exhibit higher taxonomic richness compared to those on N. tabacum stigmas (Fig. 2). Among the parasite’s stigma samples, PR MA1 and PR MA2 display similar trends and high richness, whereas PR MA3, while still rich in taxa, has a slightly lower taxa count. The host samples show lower overall richness, with NT MA1 being the least and NT MA3 the richest among them (Fig. 2).

Fig. 2
figure 2

Dendrogram of AHC (Agglomerative hierarchical clustering) results showing dissimilarity between analyzed bacterial (at the top) and fungal (at the bottom) samples of mature stigmas Phelipanche ramosa (PR MA1, PR MA2, PR MA3) and mature stigmas Nicotiana tabacum (NT MA1, NT MA2, NT MA3) (on the left); The rarefaction curves of cumulative increase of species richness (on the right).

Full size image

For fungi, the outcomes of the Bray–Curtis dissimilarity analysis, as illustrated in the AHC dendrogram, divided the stigma samples into two main clusters (Fig. 2). One distinct cluster is formed by fungi inhabiting the stigmas of N. tabacum (NT MA1), which is the least like the other samples. The second cluster consists of two groups: one includes fungi from the stigmas of P. ramosa (PR MA1 and PR MA2), while the other comprises fungi from both the host stigmas (NT MA2) and the parasite stigmas (PR MA3) (Fig. 2). Similarly to the bacterial samples, the rarefaction curves for fungi indicate that the P. ramosa samples exhibit greater taxonomic richness compared to the N. tabacum stigmas. The P. ramosa samples (PR MA1, PR MA2, PR MA3) show higher taxa richness than the N. tabacum samples (NT MA1, NT MA2), as evidenced by the higher rarefaction curve values (Fig. 2).

Moreover, the dissimilarity test using Adonis did not confirm significantly (p < 0.05) different structures of microbial communities (both for bacteria (p = 0.102) and fungi (p = 0.100)) between host and parasite stigmas.

Structure of the bacteriobiome

The STAMP analysis revealed significant differences in the relative abundance of bacterial taxa between the stigma of N. tabacum (group 1) and the parasitic plant P. ramosa (group 2) (Fig. 3A). All taxa shown exhibited statistically significant differences, with corrected q-values below p < 0.001. The tobacco stigma was enriched in taxa such as Enterobacteriaceae, Serratia, Leuconostoc, Moraxellaceae, and Acinetobacter, whereas P. ramosa stigmas showed higher proportions of Pantoea, Micrococcaceae, Pseudomonadaceae, Actinomycetales, Paenibacillus, Pseudomonas, and Exiguobacterium. Furthermore, PCA was performed to compare the bacterial communities associated with the stigmas of N. tabacum and P. ramosa (Fig. 3B). The first principal component (PC1) explained 98.1% of the total variance, clearly separating the two groups along this axis. The second and third components (PC2 and PC3) accounted for only 1.1% and 0.7% of the variance, respectively, indicating that most of the variability in bacterial community structure was captured by PC1. The distinct clustering of the two groups suggested substantial differences in the composition of their associated bacterial communities. The bacterial community composition associated with the stigmas of N. tabacum and P. ramosa differed markedly, as indicated by the heatmap of relative abundances (Fig. 3C). The P. ramosa stigma samples exhibited a higher overall abundance of several taxa, particularly Pantoea and Pseudomonas. In contrast, the tobacco stigmas were enriched in genera such as Enterobacteriaceae and Leuconostoc. These findings suggest that host plant identity, trophic strategy, and potentially the physicochemical characteristics of individual flowers may influence the structure and composition of stigma-associated bacterial communities.

Fig. 3
figure 3

Bacterial communities associated with the stigmas of Nicotiana tabacum (group 1) and Phelipanche ramosa (group 2); (A) Differences in relative abundance of bacterial taxa between group 1 and group 2 identified using STAMP analysis; (B) Principal Component Analysis (PCA) of bacterial community composition with individual samples from N. tabacum and P. ramosa stigmas; (C) Heatmap of relative abundances of dominant bacterial taxa, between the stigma of N. tabacum (group 1) and P. ramosa (group 2).

Full size image

The stigmas of P. ramosa and N. tabacum were predominantly inhabited by bacteria belonging to the phylum Proteobacteria, with a prevalence of 87.5% and 69.6%, respectively. The second most prevalent phylum was Firmicutes, which was over four times more abundant in the stigmas of N. tabacum (30.3%) compared to those of P. ramosa at 7.1%. Actinobacteria, with an average of 5.3%, were dominant in the stigmas of P. ramosa, whereas in the hosts stigmas they were present only in small amounts (0.1%). The frequency of other phyla, namely Cyanobacteria and Bacteroidetes, was below 0.02% and was recorded only in the stigmas of P. ramosa.

The analysis of stigma samples enabled us to identify bacteria from the 16 most abundant OTUs (Table 1). Notably, more OTUs were recorded in the stigmas of P. ramosa (15 OTUs) than in the host N. tabacum (11 OTUs). In the stigmas of P. ramosa, the most abundant OTU was the genus Pantoea (53.5%), whereas in N. tabacum it accounted for only 9.9%. Conversely, an inverse relationship was observed for the family Enterobacteriaceae, which constituted 38.9% of the host’s stigmas but only 7.2% of the parasite’s stigmas. Pseudomonas was dominant in the stigmas of P. ramosa (22.5%) and subdominant in those of N. tabacum (9.5%). For the genera Leuconostoc and Serratia, much higher values were recorded in the host (29.6% and 5.9%, respectively) compared to the parasite stigmas (0.3% and 0.8%, respectively). However, Pseudomonadaceae were more dominant in the parasite stigmas (2.5% vs. 0.4%). Additionally, we identified bacteria that were found only in the stigmas of P. ramosa: Paenibacillus (5.1%), Micrococcaceae (3.7%), Exiguobacterium (1.6%), Actinomycetales (1.1%), and Xanthomonadaceae (0.5%). The genus Lactococcus (0.6%) was characteristic of the host stigmas, while Moraxellaceae (2.0%) and Acinetobacter (1.9%) were more abundant in the host stigmas than in those of the parasite (0.1% each) (Table 1).

Based on the results presented in the heatmap, it was observed that all P. ramosa stigmas samples exhibited similar characteristics, while the N. tabacum stigma samples formed a distinct group (Fig. 4). Bacteria in the P. ramosa stigma samples showed the highest overall feature loads and were classified as animal parasites or symbionts, human pathogens, invertebrate parasites, and plant pathogens. The highest potential activity of physiological features in these samples was associated with aromatic compound degradation, aromatic hydrocarbon degradation, catalase positive, cellulolysis, hydrocarbon degradation, H2S production, indole production, manganese oxidation, oil bioremediation, plant hormone, plastic degradation, pyrazinamidase, siderophores, and xylanolysis. In the NT MA samples, activities were observed across 41 out of the 59 other traits (Fig. 4).

Fig. 4
figure 4

Heatmap showing the relationships between the observed traits and the bacteriobiome (on the left) and the mycobiome (on the right) of mature stigmas Phelipanche ramosa (PR MA1, PR MA2, PR MA3) and mature stigmas Nicotiana tabacum (NT MA1, NT MA2, NT MA3).

Full size image

The results of the PCA plot analysis of the first two principal components (F1 and F2), explained the range between 88.95 and 6.06% of the variance, respectively. A clear distinction is observed between the NT MA (NT MA1, NT MA2, NT MA3) and PR MA (PR MA1, PR MA2, PR MA3) groups, suggesting notable differences between them (Fig. 5). The variance (sum of squares) between the two principal components F1 and F2 is 95.01%, indicating that the PCA results confirmed statistical differences between the analyzed groups. However, the dominant contribution of F1 suggests that this component is particularly effective in differentiating the NT MA samples from the PR MA samples. Phelipanche ramosa stigma samples were correlated with Actinomycetales, Exiguobacterium, Pantoea, Micrococcaceae bacteria, and pyrazinamidase. In contrast, the NT MA samples correlated with Acinetobacter, Enterobacteriaceae, Moraxellaceae, and Lactococcus bacteria, as well as with functions such as aliphatic non methane hydrocarbon degradation, fumarate respiration, dark sulfur and dark sulfide oxidation, nitrogen fixation, mammal gut, human gut, chitinolysis, chemoheterotrophy, nitrate and nitrite ammonification, alpha-galactosidase, and CO2-fixation (Fig. 5).

Fig. 5
figure 5

Principal component analysis (PCA) showing the relationships between the observed traits and the bacteriobiome (at the top) and the mycobiome (on the bottom) of mature stigmas of Phelipanche ramosa (PR MA1, PR MA2, PR MA3) and mature stigmas Nicotiana tabacum (NT MA1, NT MA2, NT MA3); Group I: II-protease, I-ureolysis, I-aerobic chemoheterotrophy, II-aesculin hydrolysis, II-nitrate reduction to nitrite, II-urease positive, I-nitrate reduction, I-nitrous oxide denitrification, I-denitrification, I-nitrite denitrification, I-nitrate denitrification, I-dark thiosulfate oxidation, I-nitrogen respiration, I-dark oxidation of sulfur compounds, I-nitrite respiration, I-nitrate respiration, I-methylotrophy, I-methanol oxidation, I-ligninolysis, I-sulfite respiration, I-thiosulfate respiration, II-methane production, I-respiration of sulfur compounds, Gammaproteobacteria, I-dissimilatory arsenate reduction; Group II: Acinetobacter, I-aliphatic non methane hydrocarbon degradation, II-alpha-galactosidase, II-CO2-fixation, Moraxellaceae, I-fumarate respiration, I-dark sulfur oxidation, I-dark sulfide oxidation, Moraxellaceae; Group III: I-nitrogen fixation, Enterobacteriaceae, I-mammal gut, I-human gut, I-chitinolysis, I-nitrate ammonification, I-nitrite ammonification, I-chemoheterotrophy; Group IV: Pseudomonadaceae, II-indole production, I-plant pathogen, II-catalase positive, Paenibacillus, I-human pathogens all, II-plant hormone, I-animal parasites or symbionts, II- siderophores, II-H2S production, I-cellulolysis, I-aromatic hydrocarbon degradation, Pseudomonas, I-xylanolysis, I-oil bioremediation, I-manganese oxidation; Set I: I-nectar/tap_saprotroph, Alternaria, IV-sugar-rich_substrates, IV-wood,leaf/fruit/seed,soil,dung,animal_material, Cladosporium, Didymella, Candida, Trichosporon, IV-leaf/fruit/seed,roots,soil,animal_material, V-yeast; Set II: IV-leaf/fruit/seed,soil,animal_material, Powellomyces, III-algal_parasite, V-zoosporic-rhizomycelial_(chytrid-like), I-pollen_saprotroph, I-mycoparasite, IV-leaf/fruit/seed,soil,fungal_material, Trichoderma, IV-pollen; Set III: Ceratobasidiaceae, III-root-associated, Chalastospora, IV-leaf/fruit/seed, Ampelomyces, Cyphellophora, Kalmanozyma, Entoloma, I-unspecified_saprotroph, Herpotrichiellaceae, I-soil_saprotroph, II-root-associated, IV-roots,soil, Malassezia; Set IV: III-leaf/fruit/seed_pathogen, IV-soil, V-dimorphic_yeast, Dothideomycetes, Fungi, Coprinellus, Scoliciosporum, Ascomycota, Pleosporales; Set V: IV-wood, I-wood_saprotroph, Udeniomyces, V-thallus_photosynthetic, I-lichenized, I-sooty_mold, Aureobasidium, IV-leaf/fruit/seed,roots,animal_material,sugar-rich_substrates, Stereum, Phaeosphaeriaceae, Auriscalpium, Neurospora, Hormonema, II-foliar_endophyte, Pluteus, Cladonia, II-no_endophytic_capacity.

Full size image

Structure of the mycobiome

The STAMP analysis showed that all presented taxa demonstrated statistically significant differences, including highly significant results (p < 0.001) and significant results (p < 0.05) (Fig. 6A). All four identified fungal genera, Chalastospora, Chaetomium, Bensingtonia, and Kondoa, were significantly more abundant in the stigma of P. ramosa. These taxa exhibited clearly higher mean relative proportions in the parasitic plant compared to tobacco. The PCA analysis revealed a clear separation between the two groups (N. tabacum and P. ramosa), indicating distinct fungal community structures associated with each host (Fig. 6B). The first three principal components explained the majority of the variance in the dataset: PC1 accounted for 77.0% of the total variation, PC2 for 16.0%, and PC3 for 6.5%. The strong separation along PC1 suggested that host plant identity could have been the primary driver of fungal community differentiation. Minimal overlap between sample clusters further supported the presence of host-specific fungal assemblages. The PCA analysis demonstrated that fungal communities associated with the stigmas of tobacco and the parasitic plant P. ramosa were compositionally distinct. The fungal community composition differed notably between the stigmas of N. tabacum and P. ramosa, as revealed by relative abundance profiles and indicated by the heatmap of relative abundances (Fig. 6C). Notably, Chalastospora and Bensingtonia were markedly more abundant in P. ramosa stigmas, suggesting a potential preference or adaptation to the stigma environment of the parasitic host.

Fig. 6
figure 6

Fungal communities associated with the stigmas of Nicotiana tabacum (group 1) and Phelipanche ramosa (group 2); (A) Differences in relative abundance of fungal taxa between group 1 and group 2 identified using STAMP analysis; (B) Principal Component Analysis (PCA) of fungal community composition with individual samples from N. tabacum and P. ramosa stigmas; (C) Heatmap of relative abundances of dominant fungal taxa, between the stigma of N. tabacum (group 1) and P. ramosa (group 2).

Full size image

In the analyzed stigmas, Ascomycota dominated, comprising 87.2% of the fungal community in N. tabacum stigmas and 58.5% in P. ramosa stigmas. Basidiomycota were nearly three times more prevalent in the stigmas of the parasite, with 38.2% compared to 12.3% in N. tabacum. The third most common group of fungi were representatives of Mucoromycota, averaging 0.9% in P. ramosa stigmas and 0.4% in N. tabacum. Additionally, Chytridiomycota (2.4%) and Kickxellomycota (0.2%) were detected exclusively in P. ramosa stigmas, whereas Mortierellomycota (0.3%) was found only in N. tabacum stigmas. Analysis of the fungal composition on the stigmas revealed the presence of 38 of the most abundant OTUs.

A detailed analysis revealed the 38 most abundant fungal OTUs, of which 31 OTUs (89% of total abundance) were recorded in the stigmas of P. ramosa, while 27 OTUs were found in the stigmas of N. tabacum, constituting 94.3% of the total (Table 2). The genus Candida was eudominant in one of the analyzed N. tabacum samples, reaching as much as 68.6%, but it was not detected in the stigmas of P. ramosa. The P. ramosa stigmas were primarily inhabited by the genera Chalastospora, Ustilaginaceae, Pleosporales, Bensingtonia and Aureobasidium, which together constituted 64.9% of the total fungal community. Notably, the genus Bensingtonia (9.2%) was not found in the host stigmas. In N. tabacum stigmas, similar values were observed for Pleosporales (14.0% vs. 13.1%), while Aureobasidium was twice as prevalent (13.9% vs. 6.9%). However, Chalastospora and Ustilaginaceae were found in much lower amounts in N. tabacum stigmas (3.3% vs. 20.4% and 2.5% vs. 15.3%, respectively). The genera Cladosporium (4.4%) and Didymella (2.6%) were recorded only in N. tabacum stigmas. Additionally, the genera Alternaria, Trichosporon, Hanseniaspora, and Cladonia were rare or occasional in the host stigmas. However, in these two dominance classes, fungi such as Cyphellophora, Powellomyces, Ampelomyces, Kondoa, Ceratobasidiaceae, Kalmanozyma, Entoloma, Herpotrichiellaceae, Knufia and Trichoderma were characteristic only of the parasite stigmas (Table 2).

Based on the results shown on the heatmap, the clade consists of a group with progressively increasing dissimilarity, including P. ramosa stigma samples (PR MA2 and PR MA3), followed by another P. ramosa sample (PR MA1), and finally N. tabacum stigma samples (NT MA2 and NT MA1). The P. ramosa stigma samples demonstrated the highest potential activity in various functional groups: mycoparasite, plant_pathogen, pollen_saprotroph, root_endophyte_dark_septate, algal_parasite, leaf/fruit/seed,soil,animal_material, leaf/fruit/seed,soil,fungal_material, pollen, filamentous_mycelium, zoosporic-rhizomycelial_(chytrid-like) (PA MA1), soil_saprotroph, unspecified_saprotroph, root-associated, leaf/fruit/seed_pathogen, leaf/fruit/seed,animal_material, and roots, soil (PA MA3). In contrast, the N. tabacum stigma samples exhibited high activities in 17 out of 37 analysed traits. NT MA1 showed activities in groups such as animal_parasite, litter_saprotroph, nectar/tap_saprotroph, animal_material, animal_material,sugar-rich_substrates, sugar-rich_substrates, leaf/fruit/seed,roots,soil,animal_material, wood,leaf/fruit/seed,soil,dung,animal_material, and yeast. NT MA2 was correlated with lichenized, sooty_mold, wood_saprotroph, foliar_endophyte, no_endophytic_capacity, leaf/fruit/seed,roots,animal_material,sugar-rich_substrates, wood, and thallus_photosynthetic. Interestingly, these fungal features were not as active in the P. ramosa stigma samples. However, it is noteworthy that both P. ramosa and N. tabacum stigma samples showed higher activities in a few shared traits, including animal parasite, lichenized, leaf/fruit/seed pathogen, soil, dimorphic yeast, and thallus photosynthetic (Fig. 4).

The principal components account for 40.47% (F1) and 34.32% (F2) of the variance, totaling 74.79% of the overall variance (Fig. 5). The distance and distribution between the NT MA and PR MA groups suggest distinct differences in characteristics between these two sets of samples. The P. ramosa samples (PR MA1, PR MA2, and PR MA3) are closely clustered, indicating their similarity. Phelipanche ramosa stigmas were correlated with the fungi Exophiala, Knufia, Powellomyces, Trichoderma, Ustilaginaceae, as well as traits such as mycoparasite, pollen_saprotroph, root_endophyte_dark_septate, algal_parasite, pollen, leaf/fruit/seed,soil,fungal_material, leaf/fruit/seed,soil,animal_material, filamentous_mycelium, zoosporic-rhizomycelial_(chytrid-like), and Malassezia (PR MA3). In contrast, the NT MA stigma samples are more differentiated from each other. NT MA1 was related to the OTUs Alternaria, Candida, Cladosporium, Didymella, Hanseniaspora, and Trichosporon, and with traits such as litter_saprotroph, nectar/tap_saprotroph, leaf/fruit/seed,roots,soil,animal_material, animal_material,sugar-rich_substrates, wood,leaf/fruit/seed,soil,dung,animal_material, sugar-rich_substrates, animal_material,sugar-rich_substrates, and yeast. The NT MA2 sample was mainly correlated with Cladonia, Hormonema, Neurospora, Phaeosphaeriaceae, Pluteus, and foliar_endophyte, no_endophytic_capacity (Fig. 5).

Overall relationships

Network analysis illustrates the relationships between different microbes (bacteria—blue nodes, fungi—green nodes) and functional traits (violet nodes for bacteria, orange nodes for fungi) (Fig. 7). The size of the nodes is similar, indicating that each entity plays a minor role in the network. The two distinct groups stand out, each seemingly forming dense clusters with their own internal structure. The group on the left side of the analysis has more connections, especially in the middle part. These subgroups showed the relationship between several bacteria, including Gammaproteobacteria with CO2-fixation and chemoheterotrophy, Lactococcus and Moraxellaceae with fermentation, Enterobacteriaceae with nitrate ammonification, and Acinetobacter with nitrate ammonification, nitrite ammonification. The subgroup placed at the top of the analysis shows strong associations within the fungi, i.e. Stereum, Neurospora, Hormonema with wood_saprotroph, no_endophytic_capacity and wood, Aureobastidium with leaf/fruit/seed,roots,animal + material,sugar-rich_substrates and Coprinellus with leaf/fruit/seed_pathogen, and Dothideomycetes with dimorphic_yeast. Another subgroup consisting of nodes for Alternaria, Candida, Cladosporium, Didymella, and Trichosporon is clearly linked to sugar-rich_substrates, leaf/fruit/seed,roots,soil,animal_material and wood,leaf/fruit/seed,soil,dung,animal_material. The group on the right is smaller and less interconnected compared to the first one but still shows significant relationships among its nodes. A distinct subgroup is formed by Powellomyces and Trichoderma, which are associated with traits such as algal parasitism, algal_parasite, leaf/fruit/seed,soil,fungal_material, mycoparasite, pollen, pollen_saprotroph, and zoosporic-rhizomycelial_(chytrid-like). The genus Knufia is exclusively related to leaf/fruit/seed,soil,animal_material. Another significant subgroup showed the relationship between Pantoea bacteria and functional traits like plant hormone production and human pathogens. The last subgroup includes the most diverse relationships regarding the analysed features and both bacteria and fungi. The bottom-right group focuses on microbes with specialized relationships, mainly with soil and root-associated, and leaf/fruit/seed/animal_material. It highlights associations related to indole production, catalase positivity, and plant pathogenicity. These two groups represent highly distinct functional profiles, with one being more environmentally oriented and the other more biologically or pathogenically focused (Fig. 7).

Fig. 7
figure 7

Network analysis of relationships between observed properties: microbiome and physiological properties.

Full size image

Partial Least Squares Confirmatory Factor Analysis (PLS-CFA) demonstrates a moderate ability to fit the data (Goodness of fit = 0.678) (Fig. 8). A negative correlation was observed between the abundance of dominant bacterial taxa and overall bacteriobiome diversity (R = −0.947). This indicates that higher bacterial diversity is associated with a decrease in the overall presence of specific bacteria, and vice versa. Conversely, mycobiome diversity exhibits a very strong positive correlation with overall mycobiome presence (R = 0.981). This suggests that increased fungal diversity corresponds to a greater overall presence of fungi. Moreover, it was shown that there are mostly positive and strong correlated relationships between mycobiome and traits. Overall, these findings underscore the strong associations between microbiome diversity and its traits, emphasizing the importance of understanding these dynamics in relation to microbial characteristics (Fig. 7). Furthermore, after analyzing the effects of individual observed characteristics (manifest variables) in the bacteriobiome, the most abundant bacteria Pantoea and Pseudomonas were observed to be negatively correlated with Enterobacteriaceae. Within this group, negative correlations were observed between the subgroup consisting of Paenibacillus, Micrococcaceae, Pseudomonadaceae, Exiguobacterium, Actinomycetales, and Xanthomonadaceae and the subgroup consisting of Leuconostoc, Serratia, Moraxellaceae, Acinetobacter, Gammaproteobacteria, Lactococcus, and Escherichia. In turn, in the mycobiome was negatively correlated with the subgroup consisting of Candida, Aureobasidium, Cladosporium, Phaeosphaeriaceae, Didymella, Udeniomyces, Alternaria, Neurospora, Trichosporon, Hormonema, Stereum, Hanseniaspora, Auriscalpium, Cladonia, and Pluteus relative to the subgroup consisting of the remaining 23 OTUs. These negative co-associations likely reflect ecological segregation among bacterial taxa occupying similar niches on the stigmas. Such patterns may result from differences in resource use, production of inhibitory compounds, or varying environmental preferences. However, because the analysis integrated samples from both N. tabacum and P. ramosa, the observed patterns could also partially arise from host-specific effects. In this context, the negative correlations may therefore represent both ecological differentiation within microbial communities and host-driven variation in bacterial assemblages.

Fig. 8
figure 8

Partial Least Squares Confirmatory Factor Analysis (PLS-CFA) showing mutual relationships among groups of variables, including bacteriobiome, mycobiome, diversity indices, and trait categories. Bidirectional arrows indicate covariance or mutual association between latent constructs. Arrow thickness reflects the strength of the statistical association (p value): thick arrows indicate p < 0.05; mean, p = 0.1 (trend); thin, p < 0.1. Red arrows represent positive correlations; blue arrows negative ones.

Full size image

Discussion

Microbial communities can be transmitted vertically throughout the parasitic plant’s development and life cycle, as well as horizontally through environmental inputs, such as via pollinators or incidental introduction from the atmosphere or soil57. Other studies indicate that microbiota may play a key role in relationships between parasitic plants and their hosts, potentially influencing geographic and host co-specialization. Some reports also suggest the exchange of microbial taxa between parasitic plants and host plants across different locations or host species16,17. Identification of microbes colonizing a host’s tissues could be crucial for biocontrol efforts against parasitic plants16. For example, arbuscular mycorrhizal fungi, which can reduce parasitism, have been proposed as biocontrol agents for O. cumana58. It is important to note that in warmer climates, such as the Mediterranean Basin and Asia, species like P. ramosa, O. crenata, O. minor Sm., and O. cumana pose significant economic threats due to their wide host range. Although these species are not currently economically significant in Central Europe, this may change with climate warming59,60.

The relationship between stigma and seed microbiomes in parasitic plants highlights intriguing ecological and evolutionary processes, particularly regarding the overlap of specific bacterial genera. As the primary site for pollen reception, the stigma is exposed to diverse microbial populations from pollinators and the environment, which is likely to influence seed microbiome composition. The fungal communities in seeds are mostly acquired horizontally, while bacterial communities are transmitted vertically, suggesting a pathway where pollen-associated and stigma-associated microbes may enter seeds during fertilization and fruit development61. Pseudomonas, Stenotrophomonas, and Sphingomonas were widely distributed bacteria in P. ramosa seeds23 and were also identified in the stigmas of this species. Moreover, in stigmas of parasite we recorded significant amounts of Paenibacillus genus (Table 1), which constituted a minor share in P. ramosa seeds23. We identified the genus Bacillus only in the stigmas of the parasite, the presence of which in the seeds was confirmed in the work of Durlik et al.22. Among the most common fungal genera found in seeds which we identified in the stigmas of P. ramosa were Mycosphaerella and Vishniacozyma23. However, both genera represented only a minor proportion in the stigmas of P. ramosa. In contrast, the Alternaria and Cladosporium genera, which was one of the most abundant in P. ramosa seeds23, was detected exclusively in the stigmas of the host. The shared bacterial genera between stigma and seed microbiomes point to a selective mechanism retaining beneficial microbes during reproduction.

Contrasting bacterial stability and fungal diversity on parasitic plant stigmas

The highest microbial diversity (Shannon diversity (H′)) was observed in parasite stigmas, concerning fungi (2.8 vs. 2.0). Comparable results have been reported in earlier studies19,20,21. Floral communities are often characterized by lower species richness and uniformity than leaves62. In the case of parasitic plants, their leaves are strongly reduced and resemble scales, and flowers, including stigmas, may constitute a convenient microhabitat for microbes. In earlier studies, smaller bacterial OTUs were identified, indicating higher dominance compared to fungi, as more OTUs with lower abundance were observed in the samples (Tables 1, 2). This is in congruence with the results of other authors, e.g. the bacterial communities in apple stigmas were predominantly composed of Pseudomonas and Enterobacteriaceae37,38. The current results show that in the stigmas of parasitic plants, we also observe similarity in the occurrence of these groups of bacteria with notable stability across different parasite species. In contrast, fungal communities inhabiting the flower stigmas are much more varied19,20,21. The fungal endophytes in seeds were thought to be transmitted exclusively horizontally, which might explain their greater diversity in stigmas due to environmental transfer by insects, dust, rain, etc. However, recent research has revealed that some fungal endophytes can have vertical transmission rates exceeding 90% of cases63; however, this topic requires further investigation.

The lack of significant differences in microbial community structure between host and parasite stigmas, as indicated by the PERMANOVA results, may be explained by the high within-group variability (heterogeneity) observed in the Betadisper analysis. PERMANOVA is sensitive not only to differences in group centroids but also to differences in dispersion (variance) among samples. Therefore, when one or both groups show high internal heterogeneity, potential biological differences between centroids can be obscured, leading to non-significant results. This effect was evident in our bacterial data, where both groups exhibited wide dispersion and overlap in distances to their centroids. In contrast, for fungi, the apparent lack of variability likely results from the limited number of replicates (n = 2), which prevents a reliable estimation of within-group dispersion. Consequently, the non-significant PERMANOVA outcome does not necessarily indicate true similarity between communities but rather reflects differences in heterogeneity and sample size limitations.

Functional traits and ecological significance of stigma-associated microorganisms

The stigmas of both the host and parasitic plant are inhabited by microorganisms that can have potentially beneficial or pathogenic effects on the plants. Among the bacteria found in the stigmas of the parasite, Pantoea occurs as the eudominant. This widespread bacterium has a complex impact on plants, with both beneficial and harmful effects64. In the stigmas of parasitic plants, we have observed a variety of microorganisms, including potential insect and phytopathogens such as Alternaria spp. (causing black leaf spot disease), Neoascochyta spp. (responsible for leaf scorch on wheat), Phoma sp. (associated with leaf blight to root rot), Puccinia graminis (causing stem, black, and cereal rusts), and Ustilago hordei (causing covered smut of barley and oats)19,20. Beauveria bassiana, found in parasitic plant stigmas, is a pathogen known for infecting a wide range of insects. Fusarium avenaceum also has insecticidal properties and is commonly found on wheat, where it is linked to decreased crop yields. Cladosporium subuliforme is known to cause leaf spots in many plants. Some of these pathogens are particularly dangerous because they can attack plants during the pollination process. Metschnikowia cibodasensis and Microbotryum saponariae can replace pollen with fungal spores, leading to the sterilization of their host plants19. The topic of potential for pathogens to be transferred by pollinators, infecting other plants, including co-occurring native plants, is of particular concern. Flowers have been found to facilitate the spread of parasites among pollinator species, serving as key locations for disease transmission65. Additionally, infected plants might be less attractive to pollinators, as they can detect infected flowers and avoid them to protect themselves from infection66.

For the first time in research on pistil stigmas in parasitic plants19,20,21, we identified the phylum Cyanobacteria, specifically the genus Phormidium, in the stigmas of P. ramosa. Notably, this genus was absent from the host’s stigmas. Phormidium bacteria have shown potential for promoting plant growth, enhancing seed germination, growth indices, and photosynthetic efficiency in maize (Zea mays L.). These bacteria could be valuable candidates for formulating low-cost, effective biofertilizers for sustainable maize production67. Additionally, an extracellular extract (ECE) of Phormidium culture filtrate was tested for its effects on tobacco seed germination and callus differentiation. The results indicated a 40% increase in seed germination compared to the control68. Therefore, it is possible that the bacteria inhabiting the stigmas of the parasite could positively impact the life cycle of the host.

In some cases, the relationship between microorganisms and plants is not merely a one-way pathogen-plant association. For example, in tobacco stigmas, Firmicutes were found to be over four times more abundant (30.3%) compared to 7.1% in the stigmas of P. ramosa. This abundance is largely due to the genus Leuconostoc (29.6% vs. 0.3%). These bacteria can produce bacteriocins, which act as antimicrobials, helping to protect plants from harmful microorganisms69, an effect that may be important in host–parasite relationships. Interestingly, the genus Leuconostoc has also been shown to inhibit the growth of pathogens such as Allorhizobium vitis and Erwinia amylovora, which cause grapevine crown gall and fire blight70. In our study, pathogens from the genus Erwinia were not commonly observed. Additionally, these bacteria produce lactic acid, which lowers the pH in the rhizosphere, further limiting the growth of pathogenic microorganisms that could threaten the plant. A similar mechanism may occur in stigmas, where bacterial lactic acid reduces pH and decreases the likelihood of colonization by unwanted microbes. The activities of Firmicutes on plant stigmas are particularly interesting because these bacteria utilize sugars abundant in stigmas to produce lactic acid, which could influence plant development by affecting pollen hydration and germination71. Additionally, Leuconostoc can utilize complex nitrogen sources, such as those present in yeast extract, to produce lactic acid72.

Microbial influences on metabolic functioning

The metabolic functioning of stigma-associated microbiota revealed strong links between microbial diversity and biochemical activity. Bacterial taxa identified in N. tabacum stigmas, including Acinetobacter, Enterobacteriaceae, Gammaproteobacteria, Moraxellaceae, and Lactococcus, were associated with key metabolic pathways such as nitrogen fixation, chitinolysis, chemoheterotrophy, nitrate and nitrite ammonification, and CO2 fixation (Fig. 7). These processes may enhance local nutrient availability, protect reproductive tissues from microbial or insect-derived debris, and maintain a microenvironment conducive to pollination. Nitrogen-transforming bacteria can provide assimilable nitrogen to stigma tissues, while CO2-fixing taxa contribute to the stabilization of local gas exchange and the maintenance of redox balance, thereby supporting reproductive success and tissue health.

In P. ramosa stigmas, metabolic activities were dominated by bacteria such as Pantoea, Pseudomonadaceae, and Xanthomonadaceae. The abundance of Pantoea (over fivefold higher in the parasite) suggests roles in organic matter decomposition, phytohormone production, and potential pathogenic or symbiotic functions73. Members of Pseudomonadaceae and Xanthomonadaceae were linked with the degradation of aromatic compounds and hydrocarbons, cellulolysis, xylanolysis, manganese oxidation, and even plastic and oil bioremediation processes (Fig. 7). These metabolic capabilities indicate a high capacity for the breakdown of complex organic substrates, which may protect reproductive tissues from toxic accumulation or pathogen invasion74.

Fungal communities also exhibited distinct metabolic adaptations. Nicotiana tabacum stigmas were enriched in generalist fungi such as Candida, Cladosporium, Didymella, and Alternaria, known as commensals, endophytes, saprotrophs, or pathogens that influence plant health through diverse metabolic activities. Notably, Alternaria can act as a pathogen causing brown spot disease on tobacco leaves75. In contrast, the more species-rich fungal assemblage in P. ramosa stigmas may support multifunctional ecological roles, including organic matter decomposition, symbiotic root associations, and pathogenic processes that affect both parasite and host vitality (Fig. 7). The presence of yeast-like fungi is consistent with previous observations of high yeast abundance in reproductive tissues76.

Overall, the results demonstrate that even minor structural differences in microbial communities can translate into substantial functional variation. The distinct metabolic profiles of host and parasite stigmas suggest specific microbial recruitment and adaptation strategies shaped by local microenvironmental pressures and potential host–parasite microbial exchange mechanisms.

Conclusions

This study provides the first comprehensive characterization of bacterial and fungal communities inhabiting the flower stigmas of the parasitic plant P. ramosa and its host N. tabacum. The results suggest broader implications by revealing contrasting bacterial stability and fungal diversity in the stigmas, possibly reflecting distinct transmission pathways and selective pressures acting upon these microbial groups. Both stigmas were dominated by Proteobacteria, yet P. ramosa hosted more Actinobacteria (mainly Pantoea and Pseudomonas), while N. tabacum contained more Firmicutes (Leuconostoc, Enterobacteriaceae). Functionally, parasite-associated bacteria were linked with parasitism, degradation, and pathogenicity, whereas host-associated bacteria showed broader ecological functions, including nitrogen and sulfur cycling. Fungal communities in both species were dominated by Ascomycota, but P. ramosa exhibited greater structural and functional diversity, including unique Basidiomycota taxa (Chalastospora, Bensingtonia), while N. tabacum stigmas were largely colonized by Candida. These results highlight contrasting bacterial and fungal adaptations in parasite–host system and expand our understanding of stigma-associated microbiomes. They underscore the importance of these microbial dynamics in the context of host–parasite biology and open new avenues for exploring biological control strategies.