Introduction
Among the anthropogenic processes driving the contemporary loss of biodiversity and ecosystem functions, biological invasions play a major, but complicated role (Hooper et al., 2012). Their consequences for recipient ecosystems can range from devastating, to neutral, and even to positive, but they are generally difficult to predict as the effects are spread over multiple trophic levels (Simberloff, 2011). To become established in a new habitat, invaders have to face harsh conditions during transportation (e.g., unsuitable temperatures, anoxia, darkness) followed by different and sometimes novel conditions in the non-native environment. During and after transport such conditions may act as selective filters that favor certain phenotypes (Bax, Williamson, Aguero, Gonzalez, & Geeves, 2003; Blackburn et al., 2011). Compared to non-invasive populations or species, invaders tend to be more tolerant toward stressors (Walther et al., 2009), possess superior performance traits (e.g., growth rate, photosynthetic rate, nitrogen use efficiency; Van Kleunen, Weber, & Fischer, 2010), and to have higher plasticity, which enables them to express more successful phenotypes across different environments (Richards, Bossdorf, Muth, Gurevitch, & Pigliucci, 2006). Moreover, invasive plants may be morepromiscuous (or flexible ) toward microbiota, which increases the probability of acquiring the same microbial functions from a different pool of potential symbionts available in the new environment (Klock, Barrett, Thrall, & Harms, 2015; Maggia & Bousquet, 1994; Rodríguez‐Echeverría, Le Roux, Crisóstomo, & Ndlovu, 2011).
Host promiscuity (Perret, Staehelin, & Broughton, 2000) has also been termed host generalism (Rodríguez‐Echeverría et al., 2011) or microbiome flexibility (Voolstra & Ziegler, 2020) and was recently hypothesized to be a broadly occurring phenomenon among holobionts, that can promote the ability of hosts to respond and acclimate to environmental stress on short time scales (Voolstra & Ziegler, 2020). For invasive species which are transported to different environments, such a benefit would be particularly relevant as promiscuous hosts would not only be able to invade environments similar to the native environment, but would also be potentially invasive in a wider range of environments.
Whereas invasive plants have a major global impact terrestrially (Morales & Traveset, 2009; Pimentel, Zuniga, & Morrison, 2005), invasive macroalgae, or seaweeds, have caused analogous ecosystem effects in marine systems (Williams & Smith, 2007). As holobionts (see definition in Bordenstein & Theis, 2015), macroalgae are continuously interacting with microbial organisms from the water column, which colonize and penetrate their surfaces and tissues (Wahl, Goecke, Labes, Dobretsov, & Weinberger, 2012). Some of these fouling microbes may be harmless, or protective (e.g., Li et al., 2021; Saha & Weinberger, 2019), involved in the regulation of host morphogenesis (Spoerner, Wichard, Bachhuber, Stratmann, & Oertel, 2012) or spore release (Weinberger et al., 2007), or facilitating the acquisition of nitrogen and/or vitamins (Croft, Warren, & Smith, 2006; Gerard, Dunham, & Rosenberg, 1990; Kazamia et al., 2012). Other taxa may represent a threat as opportunistic or specialized pathogens (Egan, Fernandes, Kumar, Gardiner, & Thomas, 2014; Egan & Gardiner, 2016; Saha & Weinberger, 2019; Weinberger, Friedlander, & Gunkel, 1994; Weinberger, Hoppe, & Friedlander, 1997). Macroalgae manipulate the associated microbial community by means of chemical defense, for instance with specialized metabolites that target microbial fouling agents by interfering with quorum sensing (Harder, Campbell, Egan, & Steinberg, 2012). They also produce cue metabolites to attract and deter protective and harmful or (opportunistically) pathogenic symbionts (Kessler, Weiss, Kuegler, Hermes, & Wichard, 2018; Saha & Weinberger, 2019).
Recent evidence suggests that similar to plant invasions, host promiscuity may also be important in seaweed invasions. Using the invasive rhodophyte Gracilaria vermiculophylla (Ohmi) Papenfuss (synomym: Agarophyton vermiculophyllum ), Bonthond et al. (2021) conducted a common garden experiment, where native and non-native populations were subjected to a simulated invasion, experiencing disturbance followed by exposure to a new environment. Compared to native hosts, the epibiota of non-native populations changed more with respect to their epibiota in the field. The epibiota of non-native populations also became more similar to each other, suggesting that non-natives are indeed more promiscuous toward potential symbionts and may therefore acclimate more easily to new conditions. The authors also observed that epibiota associated with native holobionts dispersed more within-populations. The two measures of beta diversity between- and within-populations represent different characteristics of a community, and especially in a common garden should be interpreted separately. Whereas the beta diversity between-populations relates to the degree of change in epibiota in response to the environment, thewithin-population beta diversity is instead related to the degree of stress experienced by holobionts from that population. This stress-driven increase of variability in microbiota (or dispersion effect) is also known as the Anna Karenina Principle (Zaneveld, McMinds, & Thurber, 2017), which predicts that microbiota disperse due to a relative increase of stochastic processes (or relative decrease in deterministic processes) acting on the holobiont. Therefore, the findings in Bonthond et al. (2021) do not only suggest that non-native hosts are more promiscuous (beta diversity is lower between non-native populations) but may also indicate that non-native holobionts are less susceptible to stress or have more stable communities (beta diversity is lower within non-native populations).
Based on this idea, we aimed to specifically compare how epibiota associated with native and non-native G. vermiculophyllapopulations disperse in response to stress. A common garden experiment, similar to Bonthond et al. (2021), was conducted to simulate an invasion event by disturbing G. vermiculophylla ’s prokaryote communities with antibiotics. We also included a temperature treatment, subjecting native and non-native holobionts to optimal and (moderately) stressful thermal conditions. Here, we assumed that as subjects in the same common garden are exposed to the same environment, only processes originating from the host can vary. This implies that differences in beta diversity, which reflect differences in the ratio of deterministic and stochastic processes acting on the epibiota and therewith stability, originate from the host as well and are most likely explained by an in- or decrease of processes with which the host influences its epibiota. With two thermally different common gardens, we extended this idea to compare dispersion between two environments that only differed in temperature, expecting that non-native holobionts are less susceptible to thermal stress and therefore disperse less under stressful conditions.
Gracilaria vermiculophylla, native to the northeast Pacific, has become a widespread, invasive species along the coasts of North America, northwestern Africa, and Europe (see Krueger‐Hadfield et al., 2021, 2017 and references therein). The rhodophyte is known to be chemically well equipped to manipulate microbiota (Saha & Weinberger, 2019; Saha, Wiese, Weinberger, & Wahl, 2016), and multiple lines of evidence suggest that the interaction between host and microbes has played an important role in its successful invasion (Bonthond et al., 2020, 2021; Saha et al., 2016; S. Wang, Wang, et al., 2017; S. Wang, Weinberger, et al., 2017). Within native and non-native habitats, G. vermiculophylla covers wide latitudinal ranges, which are highly variable in temperature (Sotka et al., 2018). Not surprisingly, it is tolerant to a relatively wide temperature range. Gracilaria vermiculophylla has been found to have a thermal growth optimum ranging from 15 to 25 °C (Nejrup, Staehr, & Thomsen, 2013; Yokoya, Kakita, Obika, & Kitamura, 1999). At lower temperatures (8 °C), the alga grows slowly, but can survive for months in dark and nutrient-free conditions, which may explain its ability to survive long distance transportation (Nyberg & Wallentinus, 2009). However, qualitative observations (Weinberger et al., unpublished data) also suggested that the risk of developing disease symptoms increases at 20 °C and higher.
We collected algal thalli from native and non-native G. vermiculophylla populations and cultivated these in a common garden at an optimal growth temperature of 15 °C and an elevated temperature of 22 °C. We specifically tested the hypotheses that (i) G. vermiculophylla holobionts perform better at 15 °C compared to 22 °C (i.e., less disease symptoms and more growth) and (ii) epibiota have lower beta diversity within-populations at 15 °C compared to 22 °C . In addition, we expected that (iii) non-native algae perform better at 22 °C and (iv) their epibiota have lower within-population beta diversity this elevated temperature.