Biological invasions are considered as one of the major threats to native ecosystems’ integrity (Ehrenfeld, 2010). Invasive species, once settled, can affect ecological processes (Gandhi and Herms, 2010), engineer ecosystem structure (Crooks, 2002), or affect community dynamics (Yurkonis et al., 2005). Therefore, they can significantly alter ecosystem functioning that may result in a significant modification of ecosystem services provided. Moreover, this threat will increase with climate change (Hulme, 2017). Thus, it is crucial to understand underlying mechanisms to biological invasions to better counteract their consequences. However, most of the studies on biological invasions are performed on terrestrial environments, whereas marine environments are still poorly studied (Lowry et al., 2013; Mačić et al., 2018) despite the highlighting of numerous biological invasions (Chan and Briski, 2017; Rilov and Crooks, 2009).

Halophila stipulacea, a native seagrass from the Indian Ocean and the Red Sea (Mejia et al., 2016; Spalding et al., 2003), is considered a highly invasive species (Lowe et al., 2000; Winters et al., 2020). In the Mediterranean Sea, the first occurrences of H. stipulacea were recorded in 1894 near Greek coasts following the opening of the Suez Canal in 1869 (Winters et al., 2020). Even though this species is now found in many places, especially in eastern parts of the Mediterranean Sea (e.g., Greece, Tunisia, Libya, Syria, and Lebanon; Sghaier et al., 2011), observations point toward a relatively limited “invasion success” in this region (Nguyen et al., 2020). However, in the Caribbean Sea, where it was observed for the first time in 2002 near Grenada (Ruiz and Ballantine, 2004), H. stipulacea quickly colonized a large part of the Eastern Caribbean (e.g., Dominica, Martinique, and Virgin Islands) as well as the coasts of South America (e.g., Venezuela) in a decade (Winters et al., 2020). Moreover, it has been shown that H. stipulacea even displaces native seagrass species (e.g., Syringodium filiforme) by monopolizing their spaces (Willette and Ambrose, 2012), whereas in the Mediterranean Sea, H. stipulacea forms multispecies beds, especially with Cymodocea nodosa (Chiquillo et al., 2023). These differences in H. stipulacea invasion success between both areas is probably due to warmer conditions in the Caribbean Sea compared to the Mediterranean Sea. The actual tropicalization of the Mediterranean Sea could increase the invasiveness potential of H. stipulacea (Nguyen et al., 2020). For example, in Cap Monastir, a small patch of H. stipulacea (0.2 ha) covered more than 2 ha after only 4 years and displaced C. nodosa (Sghaier et al., 2014). Thus, this species starts to colonize northern and western parts of the Mediterranean Sea and could be an important problem in the future (Thibaut et al., 2022).

Numerous studies have highlighted what are the characteristics that make a non-native species a successful invader (Van Kleunen et al., 2010). In principle, they should present high reproductive capacity (sexual and/or asexual), wide phenotypic plasticity, high dispersal ability, and strong competitive ability. H. stipulacea presents those characteristics, namely a high tolerance to irradiance (35 to 450 μmol.m⁻².s⁻¹), to salinity levels (24 to 70 g.L⁻¹), as well as to water temperatures (17°C–42°C, Winters et al., 2020). Moreover, this species also shows a very efficient asexual reproduction by fragmentation or vegetative rhizome growth (Smulders et al., 2017) and is more fecund than larger seagrass species (Malm, 2006). Furthermore, the receiving environment should, in theory, exhibit “invadable” characteristics such as an elevated level of disturbance for the native species, availability of empty niches, a low level of biotic resistance, and a high availability of resources (Olyarnik et al., 2009). It has been thought that H. stipulacea only colonized disturbed native ecosystems. However, new evidence in the field showed that H. stipulacea displaces native species in the Caribbean Sea and, to a lesser extent, in the Mediterranean sea (Sghaier et al., 2014; Willette and Ambrose, 2012). Chiquillo et al. (2023) have demonstrated that H. stipulacea had a negative impact on S. filiforme and C. nodosa growth, whereas the presence of native species facilitates H. stipulacea settlement in a laboratory experiment. These results strongly indicate that this species is a driver of its own invasion success, implying that H. stipulacea can invade intact native ecosystems.

Allelochemicals mostly belong to specialized metabolism and act as defense compounds to suppress other plant competitors (Kong et al., 2019). Among those compounds, it has been shown that biogenic volatile organic compounds (BVOCs) are allelochemicals in terrestrial environments (Effah et al., 2019; Santonja et al., 2019; Xie et al., 2021). For instance, BVOCs emitted by Solanum lycopersicum foliage inhibited seed germination of the tropical species Amaranthus mangostanus (Kim and Kil, 2001). It has been hypothesized that BVOCs from invasive species could enhance their competitiveness since they could inhibit native species germination and/or growth, with the magnitude of the effect being regulated by co-evolution (Clavijo McCormick et al., 2023; Mollo et al., 2015). Thus, according to this hypothesis, species A would be more prepared for the biochemicals of species B if they evolved together compared to species C, which came from another biome (Callaway and Ridenour, 2004). Similar allelopathic processes, involving BVOCs, were also highlighted in marine environments (Allen et al., 2016; Sudatti et al., 2020), such as octanol showing an inhibitory effect on spore germination of Ulva prolifera (Zhang et al., 2014). Since BVOCs are lipophilic molecules, they are insoluble in water. When mediating marine species interactions, we consider that they are mainly involved in short-range or contact communication in marine environments (Mollo et al., 2014). However, they could also act on higher ranges since BVOCs can move in water through filaments over long distances (Webster and Weissburg, 2009) or their solubility can be punctually increased by water physico-chemical properties (Sander, 2023). To really decipher the action range of marine BVOCs, further investigations are required.

The aims of this study were to highlight the volatilome (i.e., all volatile compounds from a species) of the three main seagrass species from Martinique. In this area, H. stipulacea was detected in 2006 for the first time. Its presence was confirmed in 2010, and since then, the species has shown a wide distribution along the Martinique coasts at the expense of native species, mainly S. filiforme (Maréchal et al., 2013). Our hypothesis is that H. stipulacea could have a different volatilome compared to both native species, which might explain a part of its invasion success in the Caribbean area. Moreover, we expected to detect compounds from H. stipulacea having allelopathic properties, in comparison to literature, highlighted in terrestrial and marine species.

A total of 104 compounds were detected (Supplementary Files S1), with 90, 89, and 85 compounds present in H. stipulacea, S. filiforme, and T. testudinum, respectively (Figure 1A). The three species shared numerous compounds but also presented specific compounds (6 for H. stipulacea, 8 for S. filiforme, and 3 for T. testudinum). Volatilome profiles between species were quite different, with alcohol (e.g., 3-hexen-1-ol) being the major component of S. filiforme and T. testudinum (24.6 and 74.5%, respectively, Figure 1B). Then, S. filiforme volatilome included alkene (16.9%, e.g., 1-pentadecene) and aldehyde (9.6%, e.g., decanal). Concerning T. testudinum, the second major components were aldehyde (5.8%) followed by alkane (5.6%). By contrast, for H. stipulacea, major compounds of its volatilome were alkanes, representing 35% (e.g., tridecane), followed by aldehyde (15.8%) and alcohol (9.6%).

Then, to go further, PLS-DA were performed to take into account the volatilome profile as well as the compounds abundance (Figure 2). The Classification Error Rate (CER) equals to 0.14, indicated that this analyse was quite robust. PLS-DA showed that volatilomes were graphically separated from each other, suggesting that the three species had different volatilomes (in terms of compounds and abundance). This result was confirmed by PERMANOVA, indicating a significant effect of species (p-value<0.01). Then, pairwise test was performed to highlight differences between species and clearly showed that H. stipulacea volatilome was different from the two other species (p-value<0.01) whereas native species did not show any difference in their volatilome. According to VIP scores extracted from PLS-DA (1.5<VIP scores, Supplementary Files S2), 10 compounds were highlighted as the most discriminant between the three species (Table 1).

Nine of the 10 most discriminant compounds were present in higher quantities in H. stipulacea by 2 to 63-fold more. Only 3-hexen-1-ol abundance were higher in T. testudinum compared to both other species.

Our results showed, for the first time in the Caribbean area, that the invasive seagrass H. stipulacea had a volatilome significantly different from both native species, S. filiforme and T. testudinum (both showing similar volatilomes). This suggests that volatilomes do not seem to be driven by phylogeny since both species belonging to the same family showed different volatilomes (T. testudinum and H. stipulacea from the Hydrochataceae family). Similar findings have been highlighted on bryophytes (Yáñez-Serrano et al., 2024). Moreover, our results revealed the presence of several compounds in H. stipulacea volatilome, highlighted as allelochemicals in literature. For instance, within the most discriminant compounds, apocarotenoids (geranyl acetone and 6-methyl, 5-hepten-2-one also known as sulcatone) and isothiocyanates (cyclohexane, isothiocyanate) have shown strong allelopathic effects in terrestrial and marine species, as discussed below (Moreno et al., 2021; Motmainna et al., 2021).

Regarding the apocarotenoids, Jüttner (1979) showed an inhibition of cyanobacterial growth when exposed to 6-methyl, 5-hepten-2-one and even a lethal effect for geranyl acetone. Similar inhibitory effects on growth were highlighted on a green microalga, Auxenochlorella pyrenoidosa, until a concentration threshold (10mg.ml⁻¹) where no growth was observed (Ikawa et al., 2001). That could result in disturbances of both glucose uptake and respiration and/or on pigment production as shown on freshwater isolates of Chromobacterium lividum and Arthrobacter sp (Reichardt, 1981). Another type of allelochemical has been highlighted in our study, namely cyclohexane isothiocyanate. This compound belongs to isothiocyanates, well known to be produced through the degradation of glucosinolates by the enzyme myrosinase (Kliebenstein et al., 2005). It has been shown that Brassicaceae extracts, the main terrestrial plant family producing isothiocyanates (Ramirez et al., 2020), reduced germination and growth in several legume species such as Phaseolaris vulagri, Medicago sativa, and Cuscuta campestris (Almhemed and Ustuner, 2023; Choesin and Boerner, 1991; Smith, 2000). It has to be noted that other compounds detected in H. stipulacea present also allelopathic impacts (e.g., DMS, dihydroactinidiolide, and terpenoids), although they are less abundant than the previously discussed compounds. All together, these compounds could help H. stipulacea in its invasion success and to settle into a new habitat and replace native species, but this needs further investigation.

Detection of potential allelochemical is just the first step to highlight allelopathic interactions in marine invasive species such H. stipulacea. Further investigations are required to understand if and how these compounds impact competition between marine native and invasive species. That would help to better understand H. stipulacea dynamics and evaluate its invasion success under current and future environmental conditions.